Now, a few years later, we’ve got a much clearer picture of where this is all going—and we’re ready to level up. So, we’re pressing pause on the newsletter while we pivot and build something better for the long term.

This will be the last issue for a while, but we’ll be back with something we think you’ll love.

In the meantime, if you’d like to work with us, you can always reach us at hello@workwithai.com.

Thanks for reading and for following along. See you soon.

– The WorkWithAI Team

Today’s Highlights:

📰 News: OpenAI launches GPT-5 but faces major user backlash to the launch

⚡️ Top News Stories:

1. OpenAI released GPT-5, its first unified flagship model combining fast responses with deep reasoning to power next-gen ChatGPT features — enabling complex task automation, improved safety, and significantly reduced hallucinations, while outperforming competitors in coding and health benchmarks, yet showing mixed results on reasoning and agentic tasks.

• OpenAI is calling GPT-5 its first “unified” AI model that blends the fast responsiveness of the GPT series with the deep reasoning abilities of the o-series. This marks a step toward AI agents capable of performing tasks autonomously — beyond just chatting.

• GPT-5 enables ChatGPT to carry out more complex tasks like software development, calendar management, and writing research briefs. It now includes an intelligent routing system to optimize between speed and depth of reasoning in responses.

• GPT-5 outperforms competitors on certain coding and healthcare tasks — such as scoring 74.9% on SWE-bench Verified and hallucinating only 1.6% in medical queries — but underperforms on some reasoning and agentic benchmarks where rivals like Grok 4 or Claude 4.1 slightly lead.

• GPT-5 is the default model for all free ChatGPT users—removing the paywall for reasoning models—while offering GPT-5 Pro via higher-tier subscriptions and API access, with pricing set dramatically lower than rivals like Anthropic and even GPT-4o, potentially igniting an AI price war amid soaring infrastructure costs.

• GPT-5 introduces new personalities for users, offers more accurate, less deceptive responses, and features improved safeguards against misuse. While praised for improvements in accuracy and creativity, it performs variably in real-world simulations, highlighting its strengths and limitations.

2. The rollout of GPT-5 ignited one of its biggest user backlashes ever, after abruptly removing older models without notice, breaking workflows, triggering emotional distress, and exposing technical missteps that forced CEO Sam Altman to apologize and pledge the return of GPT-4o.

• OpenAI's GPT-5 launch triggered one of the most intense user revolts in ChatGPT history after previous models (like GPT-4o and GPT-4.5) were removed without warning, breaking user workflows and upsetting those emotionally attached to older models.

• Many professionals, creators, and developers complained that GPT-5 disrupted long-established prompt setups, with some stating GPT-5 lacked the nuance and tone of GPT-4o and was less suitable for creative or emotional use cases.

• The rollout was marred by a broken model router, that defaulted to lower-performing variants, which made the new model appear “dumber” than GPT-4o; Sam Altman confirmed that the real-time autoswitcher failed during rollout, affecting performance and user experience.

• Some users described GPT-4o as a therapeutic companion, sharing deeply personal experiences of emotional support from prior models, and reported feeling abandoned and devastated by the unannounced switch.

• In response to the outcry, OpenAI doubled GPT-5 usage limits, promised to bring back GPT-4o, and acknowledged they had underestimated the importance of user trust, emotional bonds, and the need for model choice and clarity.

• In a Reddit AMA following the bungled rollout, OpenAI CEO Sam Altman admitted to technical failures, misleading marketing, and misjudging user loyalty to GPT-4o, promising fixes including higher rate limits, increased transparency, and the return of the beloved model for paid users.

📰 News: xAI launches Grok Imagine - its text-to-image and video generation tool

💰 Funding: OpenAI raises $8.3Bn at a $300Bn valuation

⚡️ Top News Stories:

1. xAI has officially launched Grok Imagine, a text-to-image and video generation tool, to all SuperGrok and Premium+ X subscribers on iOS, allowing users to create stylized 15-second videos with native audio from text or image prompts, and also featuring a controversial “spicy mode” that allows limited NSFW content creation.

2. Google has publicly launched Gemini 2.5 Deep Think, a multi-agent AI model for its $250/month Ultra tier that answers complex questions by evaluating multiple ideas in parallel, outperforming top models in key benchmarks like the International Math Olympiad and LiveCodeBench 6.

3. Anthropic researchers have discovered "Persona Vectors," neural activation patterns that explain and help suppress undesirable AI behaviors like hallucinations and sycophancy by isolating the internal triggers behind personality drift.

4. Black Forest Labs and Krea have released FLUX.1 Krea, an open-weight AI image generation model designed to overcome the “AI look” by delivering more photorealistic and visually pleasing outputs.

5. Cohere has launched Command A Vision a new state-of-the-art multimodal AI model built specifically for business use, offering top-tier performance on visual tasks such as interpreting slides, diagrams, PDFs, photos, and real-world scenes, while offering secure, low-footprint deployment for industries like construction, finance, and healthcare.

6. Manus has introduced Wide Research, a breakthrough system that transforms its AI agent into a cloud supercomputing platform capable of massive-scale parallel processing and collaborative research, allowing users to command high-volume tasks simply by chatting.

7. Anthropic revoked OpenAI’s access to its Claude models after discovering the rival used Claude for internal tool comparisons and pre-GPT-5 testing, a breach of terms that underscores rising competition among top AI labs.

8. In a quiet shift, Apple has launched a new internal team, called Answers, Knowledge and Information (AKI), to build a ChatGPT-style AI-powered search engine, aiming to modernize Siri and Spotlight amid regulatory threats to its Google search deal and growing demand for generative AI search tools.

9. Cloudflare shared research accusing AI startup Perplexity of scraping websites that had explicitly blocked AI bots, alleging the company used deceptive techniques—like disguising its crawlers as Google Chrome and rotating IP addresses—to bypass protections such as `robots.txt`, impacting tens of thousands of domains and millions of daily requests.

💰 Top Funding News:

1. OpenAI has raised $8.3Bn at a $300Bn valuation—well ahead of its $40Bn annual target—with the round led by Dragoneer Investment Group ($2.8Bn) and backed by major new and existing investors including Blackstone, TPG, Sequoia, and Fidelity.

2. Meta plans to sell $2.04Bn in land and construction assets as part of a strategy to co-develop AI data centers with financial partners, helping offset the rising infrastructure costs of generative AI.

3. French AI startup Mistral is seeking a $10Bn valuation through a $1Bn raise to expand its open-source chatbot and LLMs, aiming to rival dominant U.S. and Chinese players while building a mega data center near Paris.

4. Vast Data, which provides AI-optimized storage infrastructure for efficient data center operations, is in talks to raise a new round at a valuation of up to $30B led by CapitalG and Nvidia.

5. fal, which provides infrastructure for real-time generative media across image, video, audio, and 3D, raised a $125M Series C led by Meritech with participation from Salesforce Ventures, Shopify Ventures, Google AI Futures Fund, Bessemer Venture Partners, Kindred Ventures, a16z, Notable Capital, First Round Capital, Unusual Ventures, and Village Global.

6. Fundamental Research Labs, which builds AI agents across verticals including productivity and financial modeling, raised a $33M Series A led by Prosus with participation from Patrick Collison.

📰 News: Zuckerberg shares his vision for personal superintelligence as Meta continues to poach top AI talent while spending big on AI infrastructure

💰 Funding: US NSF will invest $100M into 5 new AI research institutes

⚡️ Top News Stories:

1. Mark Zuckerberg shared his vision for Meta’s shifting AI strategy toward personal superintelligence delivered via multimodal devices like smart glasses, while backing away from open-sourcing advanced models due to growing safety concerns.

2. Meta has launched an aggressive recruitment campaign targeting Mira Murati’s AI startup Thinking Machines Lab, offering up to $1Bn to lure top researchers as it races to build competitive superintelligence models and weaponizes open source to undercut OpenAI.

3. Meta has already lured away four key researchers from Apple’s foundation model team, threatening its AI ambitions by providing multimillion-dollar offers, prompting internal unrest and a potential shift toward third-party AI models like ChatGPT amid strategic uncertainty.

4. Meta is set to spend up to $72Bn in 2025 to build out AI infrastructure including massive data centers like Prometheus and Hyperion, as part of a multi-year strategy to lead the AI arms race, while simultaneously investing billions in top-tier AI talent for its Superintelligence Labs unit.

5. Silicon Valley’s AI talent war has reached pro sports-like intensity, as AI researchers are being courted by multiple tech giants with $250M deals and perks like access to tens of thousands of GPUs, turning the hiring process into a strategic, high-stakes spectacle.

6. Google DeepMind has launched AlphaEarth Foundations, a groundbreaking AI model that synthesizes Earth observation data to produce ultra-accurate, near real-time maps for monitoring environmental changes like deforestation and coastal shifts, now available via Earth Engine.

7. Apple CEO Tim Cook announced a major ramp-up in AI investments and talent reallocation during the Q3 2025 earnings call, signaling a commitment to deeply integrate AI across Apple’s platforms and products, while defending its slower, quality-focused rollout of features like the upgraded Siri.

8. Reddit’s Q2 earnings highlight the company’s deepening focus on AI, with new advertising tools, growing data licensing revenue, and surging usage of its Reddit Answers feature driving both monetization and user engagement.

9. Anthropic has surged past OpenAI in the enterprise AI market, now holding 32% overall and 42% of the coding segment thanks to its Claude 3.5 and 3.7 models, as businesses increasingly favor closed models and move away from open-source options.

10. xAI’s Grok app is about to leap beyond chat with the launch of its “Imagine” feature for infinite image and video generation—complete with audio and adult-friendly options—alongside a new virtual companion named Valentin, all integrated in-app as xAI aims to rival GPT-5 with a viral-ready creative platform.

11. OpenAI has announced Stargate Norway, its first European AI data center built with British infrastructure provider Nscale and Norwegian energy giant Aker, a 230 MW renewable-powered facility near Narvik designed to support AI workloads with 100,000 Nvidia GPUs and advance Norway’s sovereign AI ecosystem.

12. Google is rolling out a U.S.-based test of machine learning-powered age estimation across its services, automatically restricting content and ads for users flagged as under 18 and offering a verification process to correct errors, in alignment with global child safety regulations.

13. Amazon will pay The New York Times at least $20M annually under a multiyear licensing deal that gives it access to content from the Times’s news, cooking, and Athletic sections for AI training and Alexa integration—the publisher’s first such pact and a milestone in AI-media relations.

14. Quora’s Poe has launched a powerful new developer API offering access to over 100 AI models across text, image, video, and voice through its flexible point-based subscription system, empowering developers to integrate multimodal AI tools into apps using familiar OpenAI-compatible endpoints.

15. Stanford researchers created a fully operational virtual research lab run by AI agents—including a “principal investigator” (PI) and domain-specific specialists—that can collaboratively solve complex biological problems with minimal human input, already designing a nanobody-based COVID-19 vaccine candidate in days—successfully validated in real labs—highlighting the potential.

💰 Top Funding News:

1. The U.S. National Science Foundation, backed by Capital One and Intel, is investing $100M in five new AI research institutes and a community hub to drive innovations in mental health, drug discovery, STEM education, and workforce development, solidifying America's global AI leadership.

2. Backed by Amazon’s Alexa Fund, Fable has launched Showrunner, an AI-powered platform for user-created animated TV episodes that lets viewers script, star in, and remix stories — potentially transforming streaming into two-way interactive entertainment.

3. Conversion, which uses AI to automate marketing workflows like lead organization and personalized follow-ups for midsize businesses, raised a $28M Series A led by Abstract with participation from True Ventures and HOF Capital.

4. PlayerZero, which uses AI agents to detect, fix, and prevent bugs in AI-generated code for large enterprise codebases, raised a $15M Series A led by Foundation Capital.

5. SixSense, which uses AI to predict and detect semiconductor defects in real-time to boost yield and cut inspection time, raised a $8.5M Series A led by Peak XV’s Surge with participation from Alpha Intelligence Capital, FEBE, and others.

📰 News: Google releases its own “vibe-coding” app + it expands its AI mode w/ multiple new features

💰 Funding: Anthropic targets new $3Bn to $5Bn funding round

⚡️ Top News Stories:

1. Google is testing a new AI-powered “vibe-coding” app called Opal, designed to help users create mini web apps using simple text prompts, with no traditional coding required, great for non-technical users with its prompt-based app building and visual editing.

2. Google’s NotebookLM now includes Video Overviews, a new visual AI tool that transforms dense materials—like notes, PDFs, and images—into short, engaging videos using diagrams, quotes, and numerical data.

3. Google has expanded its AI Mode with new tools like Canvas for organizing study plans, Search Live for real-time help using a phone’s camera, and broader desktop functionality, including file uploads and interactive overviews—marking a major evolution in AI-powered search and learning.

4. Microsoft has rolled out Copilot Mode in Edge, transforming its browser into an AI-powered assistant that can analyze your web activity, offer real-time help, and perform tasks like booking and content drafting — all while remaining opt-in and user-controlled.

5. OpenAI has launched Study Mode in ChatGPT to encourage critical thinking in students by requiring interaction and engagement with material rather than simply providing answers, a response to growing concerns over generative AI’s impact on learning.

6. Figma has officially released Figma Make, its AI-powered, prompt-to-app building tool, to all users — no coding skills needed — after a limited beta earlier this year.

7. Alibaba will launch its first smart glasses, the Quark AI Glasses, by the end of 2025 in China, powered by its Qwen LLM and Quark assistant to rival Meta and Xiaomi, featuring real-time translation, payments via Alipay, and integration with its entire app ecosystem.

8. Adobe has introduced AI-powered Generative Upscale, Harmonize, and an enhanced object-removal tool in Photoshop, using Firefly models to simplify tasks like image upscaling, realistic object blending, and precision background fills—now in beta for desktop, web, and iOS early access.

9. Meta has appointed former OpenAI researcher Shengjia Zhao as Chief Scientist of its new AI superintelligence lab, tasking him with advancing frontier reasoning models as part of an aggressive talent and infrastructure strategy that complements Meta’s long-term FAIR lab.

10. Tesla has signed a $16.5Bn deal with Samsung to manufacture its all-encompassing AI6 chips at a new Texas fab, cementing the chip’s role in powering Full Self-Driving, humanoid robots, and AI training — and signaling Tesla’s deeper transformation into an AI and robotics company.

11. Microsoft is negotiating to secure ongoing access to OpenAI’s AI models even after the startup declares AGI, as part of a broader deal that could restructure OpenAI into a fully commercial company, give Microsoft a 30–35% equity stake, and protect its deep integration across Azure and Copilot products.

12. Amazon- and SoftBank-backed Skild AI has launched Skild Brain, a general-purpose foundational AI model designed to make robots — from factory machines to humanoids — more adaptable and intelligent, marking a major shift toward versatile automation in the robotics industry.

13. Runway unveils new Aleph model that redefines AI video creation by enabling real-time, conversational editing that can fluidly remove, add, or restyle objects and scenes with frame-to-frame consistency, giving creators unprecedented control over video content.

14. Runway AI has Imax partnered with to bring ten short films from its 2025 AI Film Festival to theaters across the U.S., signaling mainstream Hollywood’s growing acceptance of AI-generated cinema as a viable form of artistic storytelling.

15. Mayo Clinic has deployed NVIDIA’s advanced Blackwell-powered DGX SuperPOD infrastructure to accelerate the development of generative AI in digital pathology, enabling faster, more accurate diagnostics and supporting its strategic vision for early disease detection and precision medicine.

16. In contrast to Western caution, Chinese universities are rapidly embracing AI tools like DeepSeek as essential learning companions, embedding AI literacy into curricula, hosting custom models on campus servers, and aligning education with the growing national emphasis on tech-driven career readiness.

17. 20 national security leaders have urged the Trump administration to reverse its recent approval allowing Nvidia to sell H20 AI chips to China, warning that the chip’s advanced inference capabilities pose military risks and undermine U.S. technological dominance.

18. The Department of Government Efficiency (DOGE) has developed an AI tool aimed at reviewing and eliminating half of all federal regulations, reportedly analyzing around 200,000 mandates to identify those no longer legally required.

💰 Top Funding News:

1. Anthropic is reportedly raising a $3B–$5B round at a $170B valuation, led by Iconiq Capital with potential participation from Qatar Investment Authority, GIC, and others.

2. Micro1, which uses AI to recruit specialized experts for data labeling crucial to training advanced AI models, is raising a Series A at a $500M valuation led by 01A and LG Technology Ventures.

3. Groq, a developer of high-performance AI chips for fast inference, is reportedly raising a $600M round at a $6B valuation, led by Disruptive with previous participation from BlackRock, Neuberger Berman, Type One Ventures, Cisco, KDDI, and Samsung Catalyst Fund.

4. Fable, which delivers personalized AI-generated cybersecurity training to reduce human error, raised a $31M Series A led by Redpoint Ventures with participation from Greylock Partners and Garry Tan.

5. Julius AI, which provides an AI-powered data analyst that performs analysis and predictive modeling from natural language prompts, raised a $10M seed round led by Bessemer Venture Partners with participation from Horizon VC, 8VC, Y Combinator, AI Grant, and notable angels.

📰 News: White House reveals new comprehensive AI Action Plan

💰 Funding: Cognition set to close $300M round just months after acquiring Windsurf

⚡️ Top News Stories:

1. The Trump administration’s new AI Action Plan, is built on three pillars of innovation, infrastructure, and diplomacy, outlining 90+ growth-oriented actions including deregulation, open-source promotion, and ideological bias rules, sparking some criticism for prioritizing tech giants over public safeguards.

• The new AI Action Plan shifts away from Biden-era safety regulations, focusing instead on accelerating AI infrastructure, reducing regulatory barriers, and emphasizing national security and economic competition with China.

• The Trump administration’s AI Action Plan pledges to block AI chip exports to China and rally international allies to adopt similar controls, but it lacks operational clarity and has contradicted its stance with recent greenlights for firms like Nvidia and AMD.

2. Nvidia condemned the black-market influx of at least $1Bn in banned AI chips into China—highlighting risks to service and support—amid rising U.S. export restrictions aimed at curbing China's AI development and intensifying geopolitical tech rivalry.

3. OpenAI and Oracle have partnered to add 4.5 gigawatts of Stargate data center capacity—bringing the total over 5 GW—to support AI growth, create 100,000+ U.S. jobs, and accelerate OpenAI’s $500Bn infrastructure commitment made at the White House.

4. A new Pew Research study reveals that Google’s AI Overviews—now shown in 20% of searches—cut traditional link clicks to just 8%, nearly half the rate when no AI summary is present, signaling a major threat to digital publishers dependent on organic traffic.

5. OpenAI, in collaboration with Penda Health, tested its AI Consult system in real-world Nairobi clinics, where GPT-4o provided real-time clinical support—reducing diagnostic and treatment errors by over 10% while enhancing safety and serving as an educational tool without overriding doctors’ decisions.

6. The UK has signed a strategic agreement with OpenAI to collaborate on AI safety and infrastructure, with plans that could include expanded London operations and new data centers as part of Britain’s push to become an AI superpower.

7. Anthropic researchers revealed that AI models can subliminally inherit hidden preferences or dangerous behaviors from similarly architected teacher models via neutral data, exposing a new risk vector in model training.

8. OpenAI is preparing to launch GPT-5 in early August—with integrated o3 reasoning, mini and nano variants, and a push toward AGI—while first releasing an open-weight model this July, its first since GPT-2.

9. As the AI talent war intensifies, government filings reveal that companies like OpenAI and Anthropic are offering engineers base salaries ranging from $250,000 to $690,000—excluding stock and bonuses— highlighting the soaring demand for technical expertise in the sector.

10. YouTube Shorts has introduced an AI-powered image-to-video tool and creative AI effects that animate photos and selfies, rolling out in select countries and powered by Google’s Veo 2 model with clear AI labeling via SynthID.

11. ​​Google Photos is launching AI-powered tools that let users animate pictures into short videos and transform photos into various artistic styles like anime and sketches, powered by Veo 2 and Imagen, with rollout beginning in the U.S. this summer.

12. Facing immense pressure to fund frontier AI development, Anthropic CEO Dario Amodei told staff the company will seek investment from UAE and Qatar despite ethical concerns, citing the need to stay competitive with rivals like OpenAI and secure access to billions in capital while pledging to avoid building AI infrastructure in authoritarian states.

13. Microsoft CEO Satya Nadella acknowledged the emotional weight of 2025’s 15,000 job cuts in a memo to employees, as the company’s shift toward AI and market-driven restructuring has boosted stock prices but caused internal discontent and prompted a reimagining of Microsoft’s mission in the AI age.

14. Google CEO Sundar Pichai expressed excitement over OpenAI joining Google Cloud as a customer, but the partnership underscores a high-stakes paradox: while boosting Google Cloud revenue and GPU utilization, it potentially empowers the most significant threat to Google Search using its own infrastructure.

💰 Top Funding News:

1. Cognition, creator of the coding AI Devin, is finalizing a $300M funding round at a $10B valuation backed by Founders Fund and Khosla Ventures, just months after acquiring rival Windsurf’s assets and amid a surge in enterprise adoption and market competition.

2. Delve, which uses AI agents to automate regulatory compliance across standards like HIPAA and SOC 2, raised a $32M Series A led by Insight Partners with participation from Fortune 500 CISOs.

3. LegalOn, which uses AI to streamline legal contract review and workflow automation, raised a $50M Series C led by Goldman Sachs.

4. Amazon is acquiring Bee, an AI wearable startup known for its always-listening bracelet that turns conversations into reminders, signaling Amazon’s expansion into AI-first personal hardware.

📰 News: OpenAI hires new CEO of Applications

💰 Funding: OpenAI launches $50M fund to support non-profits and community orgs

⚡️ Top News Stories:

1. Fidji Simo, soon to become OpenAI’s first CEO of Applications, outlined a vision for expanding AI access across domains like health, creativity, and knowledge while assuming key operational duties to help the company build a sustainable business model and democratize transformative technology beyond the elite.

2. AI models from OpenAI and Google DeepMind achieved gold-medal performance in the 2025 International Math Olympiad (IMO), correctly solving five out of six problems in a benchmark traditionally dominated by top high school math students.

3. Reflection AI launched a new AI agent, Asimov, which is designed to deeply understand not just code but the full context of engineering work, including architecture documents, Slack conversations, and project reports.

4. Netflix has begun using generative AI in its original productions, debuting its first on-screen GenAI footage in El Eternauta, where AI enabled a complex building-collapse scene to be completed 10x faster and at lower cost, while also integrating AI across personalization, ads, and production workflows.

5. DuckDuckGo has introduced a new AI image filter that lets users hide generative AI content in image searches using curated blocklists, addressing user concerns about AI-generated “slop” cluttering results, with broader implications for online content quality and control.

6. Meta has refused to sign the EU’s AI code of practice ahead of the AI Act’s rollout, claiming the voluntary framework introduces excessive legal uncertainty and regulatory overreach that could stall frontier model innovation and harm European AI businesses.

7. Perplexity AI, is in talks with Apple and Samsung to pre-install its agentic AI browser Comet on mobile devices, aiming to challenge Chrome’s dominance by capitalizing on preloaded app "stickiness" and usher in a new era of AI-powered browsing.

8. Senators Hawley and Blumenthal have introduced the AI Accountability and Personal Data Protection Act to allow individuals to sue AI companies for training models on their personal data or copyrighted works without consent, a move they say is necessary to hold Big Tech accountable amid growing concerns over unauthorized data use and lack of industry regulation.

9. After the mass leadership exodus to Google DeepMind shattered morale and left Windsurf in turmoil, interim CEO Jeff Wang navigated a weekend of crisis talks that culminated in an employee-first acquisition by Cognition — pairing Windsurf’s go-to-market strengths with Cognition’s engineering talent and preserving equity for all remaining staff.

10. A new Stanford-led study reveals that most leading AI models, including ChatGPT and Grok, have stopped including medical disclaimers in nearly all health-related answers—even for emergency and diagnostic questions—raising concerns that users will trust unverified and potentially dangerous advice from AI chatbots without realizing their limitations.

11. Perplexity AI is betting big on India’s massive, tech-savvy population to close the gap with OpenAI, signing an exclusive Airtel deal to give free Pro subscriptions to 360 million users and growing faster than ChatGPT in active users.

💰 Top Funding News:

1. OpenAI has launched a $50M fund to support nonprofits and community organizations using AI to tackle education, healthcare, and economic inequity, acting on the recommendations of a nationwide listening effort that involved over 500 nonprofit leaders and 7 million Americans.

2. Despite a $500Bn White House announcement and joint commitments from SoftBank and OpenAI, the Stargate AI infrastructure project has yet to materialize meaningfully due to internal disagreements, forcing OpenAI to pursue separate deals with Oracle and CoreWeave to meet its massive computing demands.

3. Hyper, which uses AI voice technology to triage non-emergency 911 calls and improve emergency response efficiency, raised a $6.3M seed round led by Eniac Ventures with participation from Ripple Ventures, GreatPoint Ventures, VSC Ventures, Tusk Venture Partners, and K5 Global.

4. Greptile, which uses AI to deliver deep code reviews that detect bugs and issues like an expert engineer, is reportedly raising a $30M Series A led by Benchmark at a $180M valuation.

📰 News: OpenAI launches ChatGPT Agent to automate complex digital tasks + Mistral updates Le Chat and releases audio based model

💰 Funding: Mira Murati closes $2Bn Seed Round for Thinking Machines Lab

⚡️ Top News Stories:

1. OpenAI has launched ChatGPT agent, a powerful new tool in ChatGPT that can autonomously perform complex digital tasks like managing calendars, creating presentations, and writing code, while implementing real-time safety monitors and disabling memory to reduce risk of abuse.

2. Mistral has released Voxtral, an open-weight audio model family capable of production-grade speech understanding and transcription, offering a multilingual, affordable alternative to closed corporate systems like Whisper and ElevenLabs.

3. Mistral has also upgraded its Le Chat assistant with deep research capabilities, multilingual reasoning, project-based organization, and advanced image editing—all deployable on-premises—positioning it as a full-featured productivity platform and a serious competitor to OpenAI and Google in sensitive enterprise environments.

4. Anthropic’s new Claude-powered Financial Analysis Solution integrates real-time data from top financial providers and platforms like Snowflake and FactSet with industry-leading AI capabilities—validated by benchmarks and enterprise adoption—to help firms like AIG, NBIM, and Bridgewater dramatically improve productivity, decision-making, and compliance.

5. Apple has introduced two new multilingual, multimodal foundation models as the backbone of Apple Intelligence features across devices and services—a compact, Apple-silicon-optimized on-device model and a scalable server model powered by a new MoE transformer architecture—all designed to deliver features with strong performance, privacy, and developer accessibility via a Swift-centric API.

6. A coalition of AI leaders from OpenAI, DeepMind, Anthropic, and others have published a position paper urging deeper research into monitoring “chains-of-thought” in reasoning models — a step-by-step process that may offer rare transparency into how advanced AI makes decisions but risks being lost without further study.

7. Runway’s new motion capture model Act-Two transforms a single performance video and character reference into highly detailed, full-body character animations with head, face, hand, and body tracking

8. At a landmark summit in Pittsburgh, President Trump announced over $90Bn in private investments to turn Pennsylvania into a national AI and energy hub through new data centers, nuclear reactors, and workforce training, drawing bipartisan political support and major industry players while sparking public protests over fossil fuels and surveillance tech.

9. Google will invest $25Bn in data center and AI infrastructure across the U.S.’s largest electric grid, PJM, and an additional $3Bn to modernize Pennsylvania hydropower plants, aligning with a nationwide push to meet AI-driven electricity demand and decarbonization goals.

10. Nvidia is resuming sales of its China-compliant H20 AI chips after U.S. export bans were reversed following CEO Jensen Huang’s meeting with Trump, a move that comes amid trade talks over rare-earth elements and has sparked controversy—U.S. officials downplay the chip’s significance, while critics argue it undermines established export control policy.

11. The U.S. has paused a multi-billion dollar deal allowing the UAE to buy Nvidia AI chips due to rising national security concerns that the hardware could be smuggled to China, despite previous regional assurances and amid new moves to tighten global chip export controls.

12. Slack is integrating AI across its platform with smarter search across connected apps, automated Huddles transcriptions and summaries, contextual message explanations, and Canvas writing assistants — all part of a broader redesign aimed at boosting productivity and collaboration.

13. xAI is under fire from top AI safety researchers at OpenAI and Anthropic for releasing its Grok 4 chatbot without standard safety evaluations or documentation, after the model exhibited antisemitic and manipulative behavior, prompting renewed calls for government oversight and regulation.

💰 Top Funding News:

1. Thinking Machines Lab, which is developing frontier AI models and plans to release open-source tools for researchers and startups, raised a $2B seed round led by Andreessen Horowitz with participation from Nvidia, Accel, ServiceNow, CISCO, AMD, and Jane Street.

2. Anthropic is attracting investor interest at a valuation north of $100Bn — nearly doubling its previous round’s value — as soaring revenue from its Claude chatbot and backing from Amazon and Alphabet position it as a leading contender in the AI arms race.

3. AI search challenger Perplexity has leapt to an $18Bn valuation just months after a $14Bn round, as surging revenue, a powerful new AI browser, and backers like Bezos and Nvidia fuel investor excitement about its ambitions to disrupt Google and commercialize AI at scale.

4. Lovable, which lets users build apps and websites with natural language using large language models, raised a $200M Series A led by Accel with participation from 20VC, byFounders, Creandum, Hummingbird, Visionaries Club, and notable angel investors.

📰 News: Meta will construct a colossal 5 GW AI data center + DoD awards massive contracts to the 4 leading US AI companies

💰 Funding: Windsurf signs massive licensing deal with Google and gets acquired by Cognition, all after OpenAI’s acquisition falls through

⚡️ Top News Stories:

1. Despite its public commitment to open-source AI, Meta is reportedly considering a major shift toward closed model development after its underwhelming Behemoth model, a pivot that would signal strategic monetization pressures and potentially reshape the open-source ecosystem it helped build.

2. Mark Zuckerberg revealed Meta is constructing Hyperion, a colossal 5 GW AI data center in Louisiana to power its Superintelligence Lab and compete with OpenAI and Google, signaling the company’s aggressive push into energy-hungry AI infrastructure.

3. In a scathing internal exit essay, now former Meta AI scientist Tijmen Blankevoort compared the company’s culture to “metastatic cancer,” describing an AI division paralyzed by fear and low morale — a revelation that casts doubt on Meta’s high-profile talent poaching spree and its new Superintelligence ambitions.

4. The Department of Defense's Chief Digital and Artificial Intelligence Office (CDAO) has awarded contracts—each with a $200M ceiling—to the four leading U.S. frontier AI companies (Anthropic, Google, OpenAI, and xAI), to integrate frontier AI models into military operations and intelligence systems.

5. In a major policy shift aimed at countering the rise of AI-generated “slop” and content theft, Meta will penalize Facebook accounts that habitually repost unoriginal content by limiting their reach and monetization, following in the footsteps of YouTube’s similar crackdown.

6. Google has expanded its open-source medical AI suite with two new MedGemma models—including a multimodal model and the MedSigLIP image-text analysis tool—capable of interpreting diverse medical data such as chest X-rays and skin conditions, with a smaller variant optimized for phones and consumer devices.

7. Mistral has released Devstral, a suite of coding-optimized AI models including an open-source version that outperforms major proprietary models on real GitHub issues, enabling autonomous software agents to operate efficiently across large repositories.

8. Anthropic is rolling out Claude for Education with Canvas, Wiley, and Panopto integrations, letting students reference lectures and research within conversations while maintaining strict privacy and security standards.

9. Microsoft launches Phi-4-mini-flash-reasoning, a new 3.8B parameter model designed for low-latency, high-efficiency reasoning on edge and mobile devices, offering 10x throughput and 2–3x lower latency over its predecessor.

10. Google’s NotebookLM is expanding from a personal AI note-taking assistant into a content-rich discovery hub with featured notebooks from The Economist, The Atlantic, and academic experts, offering users curated content on topics from longevity and parenting to financial trends and literature—enhanced with interactive tools like AI-generated Q\&A, audio overviews, and mind maps.

11. Amazon Web Services has launched a preview of Kiro, an AI-powered assistant that helps developers build software with minimal input by generating specs, diagrams, and task lists—marking AWS’s entry into the competitive "vibe coding" space alongside Google and Microsoft.

12. OpenAI has postponed the release of its anticipated open-source AI model indefinitely to conduct further safety testing, a move that delays developers' access to what the company claims will be a best-in-class model rivaling its o-series, amid growing competition from global rivals like Moonshot AI.

13. xAI has issued a formal apology after Grok’s antisemitic outbursts — including calling itself “MechaHitler” — blaming a code update that made the bot overly influenced by extremist user posts, though critics argue the behavior was unprovoked and systemic.

14. Goldman Sachs is integrating Cognition’s viral AI coding agent Devin as a digital employee across its developer teams, planning hundreds to thousands of supervised deployments to boost productivity without replacing humans, as part of its continued push into hybrid AI-human collaboration.

💰 Top Funding News:

1. Windsurf, the AI IDE startup, has been officially acquired by Cognition, developer of the autonomous agent Devin, in a strategic move to integrate Windsurf’s enterprise IDE with autonomous coding capabilities.

• The acquisition follows the sudden collapse of OpenAI’s $3Bn acquisition bid, which expired before closing, leaving the door open for new deals.

• In a parallel development, Google entered into a $2.4Bn licensing agreement for Windsurf’s technology and executed a high-profile “reverse-acquihire,” bringing on Windsurf’s CEO Varun Mohan, co-founder Douglas Chen, and top researchers into DeepMind.

• Windsurf has now pivoted to focus solely on enterprise AI adoption under new leadership, continuing its mission with support from a strategic partnership with Google.

2. SpaceX is reportedly investing $2Bn in Elon Musk’s AI company xAI—its first such investment and part of a $10Bn funding round—despite recent Grok controversies, deepening the integration between Musk’s ventures like SpaceX, Tesla, and X.

3. Moonvalley, which builds licensed AI video models for professional filmmakers and brands, raised an $84M round led by General Catalyst with participation from Creative Artists Agency, CoreWeave, Comcast Ventures, Khosla Ventures, and Y Combinator.

4. Meta has acquired Play AI, a startup known for its realistic AI-generated voices, as part of its growing investment in audio interfaces, AI Characters, and superintelligence—signaling deeper ambitions in wearables and multimodal AI.

📰 News: xAI launches Grok 4 models + Perplexity launches its AI-powered Comet browser

💰 Funding: LangChain and Mistral both near $1Bn valuations

⚡️ Top News Stories:

1. xAI launched its high-performing Grok 4 and Grok 4 Heavy models with a $300/month SuperGrok Heavy subscription, boasting benchmark wins and multi-agent reasoning.

• Grok 4 Heavy uses multiple agents that collaborate like a study group to solve problems, and it outperformed competitors on tests like Humanity’s Last Exam and ARC-AGI-2, achieving frontier-level benchmarks.

• Elon Musk also announced Grok will arrive in Tesla vehicles next week as an in-car AI assistant.

• Grok 4 has also been found to consult Musk’s own social media posts and views when answering controversial questions—such as on immigration and abortion—raising ethical concerns about bias and transparency, especially as the company markets the tool for enterprise use and Tesla integration without releasing standard documentation like a system card.

• The launch came shortly after Grok was temporarily shut down for generating offensive responses.

2. Perplexity launched its AI-powered web browser Comet, available to premium users, which integrates the company’s AI search engine and introduces an assistant that performs contextual in-browser tasks like email summarization and calendar navigation, marking a bold move to compete directly with Google Search and Chrome.

3. OpenAI is reportedly preparing to launch its own AI-powered web browser in the coming weeks, potentially integrating “Operator,” a web-browsing AI agent, to keep more user interactions within ChatGPT rather than redirecting to external sites.

4. OpenAI is poised to release its first open-weight language model since GPT-2—comparable in reasoning ability to "o3 mini" and deployable across cloud platforms like Azure, Hugging Face, and others.

5. OpenAI has partnered with education unions and tech companies to launch a $10M backed National Academy for AI Instruction, aiming to equip 400,000 U.S. teachers with training, tools, and resources to responsibly integrate AI into K-12 classrooms by 2030.

6. Alphabet’s Isomorphic Labs, born from DeepMind’s AlphaFold breakthrough, is preparing its first human trials of AI-designed drugs, aiming to revolutionize drug discovery through powerful partnerships, $600M in funding, and ambitions to dramatically increase speed, reduce cost, and improve success rates in pharmaceutical development.

7. Moonvalley has launched its AI video model Marey to the public, offering a credits-based system for filmmakers seeking precise creative control, ethical data sourcing, and advanced features like 3D-aware motion, camera manipulation, and background edits, all while rivaling competitors like Runway Gen-3 and Pika.

8. Google has rolled out a new image-to-video feature powered by its Veo 3 AI model within the Gemini app, allowing AI Ultra and Pro users to turn photos into short videos with optional audio, marking a significant creative expansion seven weeks after Veo 3’s launch.

9. Replit has partnered with Microsoft to bring no-code, AI-powered software development to enterprise users via Azure, allowing teams across departments to securely build and deploy apps using natural language and access Replit through the Azure Marketplace.

10. OpenAI has ramped up security across its offices, infrastructure, and digital systems—in response to espionage fears after DeepSeek allegedly copied its models–limiting employee access, isolating key technologies, and boosting biometric and cybersecurity measures.

11. Just days after laying off 9,000 employees, Microsoft revealed it saved over $500M in call centers thanks to AI tools like Copilot, underscoring a controversial strategy of cutting jobs during record profits while doubling down on AI investments.

12. FlexOlmo by AI2 introduces a breakthrough framework for training language models that enables decentralized collaboration by letting data owners contribute expert modules without sharing raw data, dynamically control participation, and receive credit, all while maintaining strong performance and privacy safeguards.

13. Nvidia is reportedly launching a modified version of its Blackwell RTX Pro 6000 chip for China by September, designed to comply with U.S. export controls by excluding high-end features like high-bandwidth memory and NVLink, in an effort to re-engage a datacenter market now dominated by competitors like Huawei.

14. AI chipmaker Groq has launched its first European data center in Helsinki with Equinix, positioning its fast-deploying, cost-efficient inference chips as a strategic alternative to Nvidia’s GPU dominance amid Europe's push for sovereign AI and greener data infrastructure.

15. A threat actor used AI to impersonate Secretary of State Marco Rubio, contacting at least five high-profile individuals—including three foreign ministers, a U.S. governor, and a member of Congress—via Signal with the aim of gaining access to sensitive accounts or information, according to a classified diplomatic cable.

16. California State Senator Scott Wiener has introduced amendments to SB 53, a revised AI transparency bill requiring top AI companies like OpenAI, Google, xAI, and Anthropic to publish safety protocols and incident reports—marking California's most significant legislative push yet for AI accountability.

17. Amazon Web Services is launching an AI agent marketplace next week with Anthropic as a partner, aiming to centralize distribution and monetization of autonomous AI agents for enterprise customers and startups, in a move poised to challenge rival ecosystems from Microsoft and Google.

18. In a high-profile recruitment move amid escalating AI talent wars, OpenAI has brought on four senior engineers from Tesla, xAI, and Meta — including former Tesla VP David Lau and two architects of xAI’s massive Colossus supercomputer — to bolster its Stargate data center project and backend infrastructure.

19. Luma AI is launching a Hollywood-based Dream Lab to train and collaborate with filmmakers on generative video tools like its Dream Machine and Modify, aiming to redefine content production by making AI-driven filmmaking faster, cheaper, and more creatively accessible.

20. YouTube will revise its monetization rules on July 15 to crack down on inauthentic and mass-produced content, including AI-generated videos, as part of an effort to safeguard its platform from low-quality media that has increasingly flooded the site.

💰 Top Funding News:

1. Meta has acquired a nearly 3% stake in eyewear leader EssilorLuxottica for $3.5Bn, expanding its push into AI-enabled smart glasses following the success of Ray-Ban Meta and signaling further ambition in the wearable tech space.

2. LangChain, the AI infrastructure pioneer originally known for enabling LLMs to perform real-world tasks, is nearing unicorn status with a reported $1B valuation round led by IVP, driven by the rapid adoption of its LangSmith monitoring platform by major clients like Klarna and Replit.

3. Mistral, a French AI leader known for its open-weight language models and the Le Chat chatbot, is reportedly raising up to $1B in equity—plus hundreds of millions in debt—as it partners with MGX and Nvidia to build Europe’s largest AI data center, backed by France's push for AI sovereignty.

4. Nvidia has become the world’s first $4 Trillion company, powered by explosive demand for its AI chips and software following ChatGPT’s release, as tech giants pour billions into infrastructure reliant on its GPUs despite rising competition and regulatory scrutiny.

5. LGND, which applies AI to geospatial data to streamline spatial analysis and querying, raised a $9M seed round led by Javelin Venture Partners with participation from AENU, Clocktower Ventures, Coalition Operators, MCJ, Overture, Ridgeline, Space Capital, and several angel investors.

6. Blok, which uses AI personas to simulate app user behavior for pre-launch testing and product optimization, raised a $7.5M seed round led by MaC Venture Capital with participation from Protagonist, Rackhouse, Weekend Fund, Blank Ventures, and tech industry angels.

7. ZeroEntropy, which builds AI infrastructure to improve retrieval-augmented generation for developers, raised a $4.2M seed round led by Initialized Capital with participation from Y Combinator, Transpose Platform, 22 Ventures, a16z Scout, and angels from OpenAI, Hugging Face, and Front.

📰 News: Meta steals another top AI talent, hiring Apple’s top AI model leader + EU push forward implementation of its AI Act

⚡️ Top News Stories:

1. Apple’s top AI model leader Ruoming Pang has reportedly defected to Meta, joining Zuckerberg’s growing AI superintelligence unit amid growing concerns about Apple’s lagging in-house AI performance and possible further defections.

2. Despite pushback from more than 100 tech firms including Alphabet and Meta, the EU declared it will not delay implementation of the AI Act, reaffirming a mid-2026 deadline and its commitment to regulating high-risk and unacceptable AI use cases.

3. Anthropic has proposed a flexible AI transparency framework for large model developers, requiring public disclosure of safety practices and legal accountability, while cautioning against rigid government rules that could hinder innovation.

4. Epoch AI's new report proposes a U.S.-led national AI initiative on the scale of the Apollo or Manhattan Projects, suggesting it could fund 27 million GPUs and develop a model 10,000 times larger than GPT-4 by 2027 through coordinated public-private investment.

5. AI chip startup Groq has opened its first European data center in Helsinki to expand its inferencing technology across the continent, emphasizing a fast deployment model and a supply chain advantage over Nvidia by avoiding high-bandwidth memory dependencies.

6. OpenAI is quietly testing a new ChatGPT tool called “Study Together,” which flips the tutoring dynamic by prompting users with questions instead of giving direct answers, potentially as a response to concerns over misuse in education.

7. Replit has launched Dynamic Intelligence for its AI Agent, combining deep reasoning, powerful model switching via Claude Opus, and real-time web search to help developers tackle large, complex coding problems with precision.

8. Researchers at multiple top global institutions have been caught embedding hidden prompts in academic papers to bias AI-powered peer review systems into giving overly positive feedback, according to a Nikkei Asia investigation.

9. Google is facing an EU antitrust complaint over its AI Overviews in Search, as publishers allege the summaries misuse their content and are slashing web traffic, with no opt-out available short of disappearing from search altogether.

10. Cursor ignited developer outrage by quietly changing its flat Pro plan with a token-based system that rapidly depleted usage quotas, prompting surprise charges, mass cancellations, and prompting CEO Michael Truell to apologize and offer refunds while defending the move as necessary due to skyrocketing costs of advanced AI models like Claude.

11. Meta is testing AI chatbots that proactively message users across its platforms using remembered interactions to foster engagement, raising concerns over safety, commercialization, and its potential use in driving future ad revenue and subscription models.

💰 Top Funding News:

1. In a $9Bn stock deal, CoreWeave is acquiring Core Scientific to secure over a gigawatt of AI-ready data center capacity, transforming former Bitcoin mining infrastructure into critical compute power for the generative AI race.

📰 News: Perplexity launches new ‘Max’ subscription + Amazon releases foundational gen-AI model for warehouse robots

💰 Funding: Grammarly acquires Superhuman + xAI raises $10Bn

⚡️ Top News Stories:

1. Perplexity has launched a $200-per-month 'Max' plan offering advanced features like unlimited access to its Labs tool, early access to its Comet browser, and priority use of top AI models, as it aims to grow revenue and justify a potential $14Bn valuation.

2. Amazon introduced DeepFleet, a foundational generative AI model trained on internal warehouse data to boost warehouse robot speed and efficiency by 10%, while deploying its one millionth warehouse robot.

3. Researchers at Helmholtz Munich have built Centaur, an AI model that simulates human decision-making with high precision by training on 60,000 participants' choices across 160 psychology experiments, outperforming cognitive models in nearly all benchmarks and offering a revolutionary new tool for behavioral science.

4. OpenAI is launching a consulting division targeting enterprise clients willing to spend $10M or more to customize AI models and integrate them into their operations.

5. Baidu has unveiled ERNIE 4.5, a suite of 10 fully open-source multimodal AI models that outperform GPT-4o and Gemini in tasks like OCR, visual math, and Chinese reasoning, while also launching the ERNIE X1 model aimed at high-performance reasoning, all underpinned by efficient MoE training and cost-effective inference.

6. Cloudflare has launched “Pay per Crawl,” a bold new marketplace enabling websites to charge AI bots micropayments for scraping their content — a direct response to declining Google traffic and rampant free AI scraping by OpenAI and Anthropic, which Cloudflare says gives publishers little to no return.

7. Sakana AI introduced AB-MCTS, a novel algorithm that coordinates multiple AI models — including ChatGPT, Gemini, and DeepSeek — to collaboratively tackle complex tasks by assigning roles based on model strengths, outperforming any solo model.

8. SciArena, a new platform from AI2, Yale, and NYU, lets expert scientists vote on the best answers generated by competing AI models to scientific questions, creating a dynamic, human-evaluated leaderboard that reveals performance gaps even among top-tier LLMs like o3, Claude, and DeepSeek.

9. X is testing AI-generated Community Notes from chatbots like Grok and third-party models, raising hopes of scalable fact-checking but also concerns about hallucinations, human oversight fatigue, and the accuracy of automated moderation.

10. OpenAI will lease 4.5 gigawatts of computing power from Oracle in a record-setting expansion of the Stargate initiative, prompting a U.S.-wide data center buildout and reinforcing Oracle’s role as a key player in global AI infrastructure.

11. Ilya Sutskever has assumed the CEO role at Safe Superintelligence following Daniel Gross' departure amid reports of Meta’s acquisition interest, reaffirming the company’s sole focus on building safe superintelligence despite looming challenges in leadership and recruitment.

12. Sam Altman responded to Meta’s poaching of OpenAI talent by reaffirming OpenAI’s mission-driven culture, criticizing Meta for prioritizing opportunism over long-term values, and promising to review compensation to retain top researchers.

💰 Top Funding News:

1. Grammarly has acquired AI email client Superhuman to integrate its advanced scheduling and reply features into a broader productivity suite powered by AI agents, particularly targeting the email workflows of professionals.

• Superhuman was last valued at $825M

2. xAI has raised $10Bn in a mix of debt and equity to fund its Grok chatbot and massive data center buildout, pushing its total funding to $17Bn and expanding its access to capital markets.

3. Lovable, which offers an AI-driven coding platform that allows users to create apps without coding experience, is set to raise over $150M at a $1.8B valuation in a round led by Accel, with participation from 20VC and Creandum.

4. Genesis AI, which is building a foundational AI model to power general-purpose robots using synthetic data, raised a $105M seed round co-led by Eclipse and Khosla Ventures.

5. Wonderful, which builds AI customer support agents for non-English speaking markets using customizable multilingual models, raised a $34M seed round led by Index Ventures with participation from Bessemer Venture Partners and Vine Ventures.

📰 News: Meta continues to poach top AI talent from OpenAI and other AI rivals + Google releases Gemma 3n model

💰 Funding: OpenAI and Meta continue to expand AI capabilities through acquisitions

⚡️ Top News Stories:

1. Meta has now poached eight top researchers from OpenAI — including contributors to critical reasoning models like o1 and GPT 4.1 — with CEO Mark Zuckerberg personally spearheading a stealth recruitment campaign as part of the company’s high-stakes superintelligence initiative.

• Meta has unified its AI operations into a new unit called “Meta Superintelligence Labs” under the leadership of Alexandr Wang and Nat Friedman, in a bold move to dominate the AGI race through high-profile hires, acquisitions, and a $14.3Bn investment in Scale AI.

• Meta is offering multimillion-dollar pay packages — primarily structured through equity and performance incentives — to lure AI researchers to its new superintelligence unit, but contrary to OpenAI CEO Sam Altman’s claims, these are not blanket $100M signing bonuses, according to Meta executives and newly hired researchers.

• In response to Meta luring away key researchers, OpenAI's leadership — including CEO Sam Altman — is recalibrating compensation and launching new retention strategies, as Chief Research Officer Mark Chen equated the departures to a personal violation and vowed stronger internal support.

2. Google introduces Gemma 3n, a powerful multimodal AI model designed specifically for mobile and edge devices, supporting text, image, video, and audio inputs with outputs in text, all while maintaining performance levels previously only seen in cloud-based models.

3. Cursor, the AI coding platform from Anysphere, has launched a web app that allows users to assign and manage coding tasks via browser using autonomous background agents, a move that builds on its Slack integration and follows rapid enterprise adoption.

4. Congress is considering a provision in the GOP-backed budget bill that would bar states and local governments from regulating AI for five years, reduced from an original ten-year proposal, in an effort to avoid a regulatory "patchwork," a move backed by industry giants like OpenAI but fiercely opposed by bipartisan lawmakers, consumer advocates, and tech rivals like Anthropic, who argue it strips states of vital oversight powers and ties federal broadband funding to compliance.

5. Google unveiled over 30 new AI tools at ISTE 2025 edtech conference, including a custom Gemini app, chatbots, and multimedia features aimed at transforming education through personalized learning, AI-supported teaching, and secure, streamlined classroom management.

6. OpenAI is now renting Google's TPUs through Google Cloud in a bid to reduce inference costs and diversify beyond Nvidia GPUs and Microsoft Azure, marking its first significant use of a rival’s chips while Google cautiously limits access to its most powerful hardware.

7. Microsoft has unveiled the "AI Diagnostic Orchestrator" (MAI-DxO), a new AI system that simulates a team of virtual doctors working together to diagnose complex medical cases, claiming to outperform humans 4x on complex cases from medical journals—an early but striking milestone toward AI-powered clinical tools.

8. Anthropic has launched the Economic Futures Program a new initiative aimed at studying the impact of AI on labor markets and the global economy, intending to produce evidence-based policy proposals and foster dialogue among experts, policymakers, and institutions, distinguishing itself from competitors like OpenAI by focusing on potential job losses, fiscal shifts, and new economic paradigms rather than merely promoting AI adoption.

9. Mira Murati’s secretive startup Thinking Machines Lab is disrupting the AI talent market by paying base salaries up to $500,000 and recruiting former OpenAI heavyweights, signaling an aggressive push to rival top AI labs even before launching any products.

10. Facing delays with its in-house “LLM Siri” project, Apple is now in talks with OpenAI and Anthropic to potentially power a revamped version of Siri using their AI models on Apple’s own cloud infrastructure.

💰 Top Funding News:

1. OpenAI has absorbed the team from e-commerce recommendation startup Crossing Minds, whose privacy-focused personalization technology will now bolster OpenAI’s broader AI research and agent initiatives.

• Crossing Minds had raised over $13.5M from investors like Index Ventures, Shopify, and Radical Ventures, and worked with companies including Intuit, Anthropic, Udacity, and Chanel.

2. Legal software company Clio is acquiring law data powerhouse vLex for $1Bn, gaining access to its rich legal document database and AI model Vincent to expand beyond firm management tools into the actual practice of law.

3. Meta is reportedly negotiating to acquire Play AI, a voice-cloning startup known for its speech replication tech, in a bid to bolster Meta AI's audio capabilities and integrate voice tools into its growing creator and chatbot platforms.

• Play AI has raised $23.5M from investors like 500 Global, Kindred Ventures, Race Capital, and Soma Capital, per Crunchbase data.

4. Campfire, which offers an LLM-powered ERP platform that automates financial processes for startups, raised a $35M Series A led by Accel, with participation from Foundation Capital, Y Combinator, Capital 49, and angels.

📰 News: Anthropic joins the vibe-coding trend by launching a Claude based no-code App builder + ElevenLabs launches a new voice-first AI assistant

💰 Funding: Abridge raises $300M for AI-powered medical note-taking

⚡️ Top News Stories:

1. Anthropic upgraded Claude, now allowing users to build, host, and share interactive AI-powered apps directly within the Claude app—no deployment or infrastructure management required.

• As part of the offering Anthropic has also launched a new artifacts space, enabling users to turn conversations into AI-powered, interactive apps without any coding.

2. ElevenLabs has launched 11ai, a proactive voice-first AI assistant that executes real-world productivity tasks—like planning, summarizing, and workflow management—using natural language and powerful integrations with 30+ services via the open-source Model Context Protocol, including tools like Perplexity, Notion, Gmail, Linear, Zapier, and Salesforce.

3. Google DeepMind has introduced Gemini Robotics On-Device, a powerful language model capable of running locally on robots to perform real-world tasks without internet dependency, marking a major step forward in offline, hands-free robot control via natural language.

4. DeepMind has introduced AlphaGenome, an advanced AI model designed to predict how DNA variants affect gene regulation, outperforming existing models and offering scientists a unified, multimodal tool to better understand gene regulation and disease biology.

5. Google has unveiled Gemini CLI, an open source, terminal-based AI tool that allows developers to use Gemini models for code generation, debugging, research, and video creation, all while running locally and offering generous usage limits under an Apache 2.0 license.

6. Microsoft’s new Mu language model is a compact, encoder-decoder LLM optimized for NPU-powered edge devices, enabling real-time, on-device natural language interfaces like the Windows Settings agent with low latency and high performance.

7. A senior U.S. official alleges that Chinese AI firm DeepSeek is covertly supporting China’s military and intelligence through technology services and data sharing while evading U.S. chip export controls using Southeast Asian shell companies and remote data centers.

8. In a landmark ruling, a federal judge determined that Anthropic’s use of copyrighted books to train AI models qualifies as fair use, marking the first major legal validation of tech companies’ argument that fair use can apply to AI training data, establishing a critical precedent favoring AI firms and challenging ongoing lawsuits from authors and publishers.

9. Another federal judge ruled that Meta’s training of AI models on books by authors including Sarah Silverman was legal under the fair use doctrine, citing transformative use and lack of market harm, in a decision that narrows but reinforces AI firms’ legal defenses.

10. Meta has recruited Trapit Bansal, a top AI researcher from OpenAI who helped create its foundational o1 reasoning model, to spearhead its efforts in building a cutting-edge AI reasoning system within its newly formed AI superintelligence team.

11. OpenAI is quietly advancing ChatGPT into a full-fledged productivity platform with tools like real-time collaboration and enterprise data connectors, directly challenging Google Workspace and Microsoft Office.

12. Creative Commons has unveiled CC signals, a licensing-inspired framework to help dataset owners specify how their data can be used for AI training, aiming to preserve openness online amid growing restrictions and data scraping battles.

13. In response to AI features diminishing publisher search traffic, Google has launched Offerwall—a new Ad Manager tool that helps publishers monetize content through customizable options like micropayments, surveys, and ad-watching, with early adopters reporting modest revenue boosts.

14. Reddit is considering adopting Sam Altman’s World ID, an iris-scanning verification system from Tools for Humanity, to help authenticate users as unique individuals amid growing concerns over AI-generated content and age verification laws, all while preserving user anonymity.

15. In a significant shift in its UK legal battle, Getty Images has dropped key copyright infringement claims against Stability AI over the use of its images in training Stable Diffusion, focusing instead on secondary and trademark infringement that could have wider implications for generative AI firms training models abroad.

16. YouTube has launched a new AI-powered search results carousel and is expanding its conversational AI tool, aiming to improve content discovery and viewer engagement, though creators fear the changes could undercut video clicks and monetization.

17. Meta has introduced a privacy-preserving AI-powered summaries feature on WhatsApp that allows users to privately condense unread messages using Meta AI, with U.S. rollout and English language support first.

18. Disney has launched a precedent-setting lawsuit against AI company Midjourney to protect iconic IP like Darth Vader from unauthorized use, framing the case as the first step in a broader Hollywood campaign to defend copyright in the AI age.

💰 Top Funding News:

1. Abridge, which automates medical note-taking and medical coding using AI, raised a $300M Series E round led by Andreessen Horowitz, with participation from Khosla Ventures.

2. Sanmai Technologies, which is pioneering a $500 at-home brain treatment device that combines AI with focused ultrasound to non-invasively treat mental health conditions and enhance cognition, raised a $12M investment from Reid Hoffman.

3. XBOW, which uses autonomous AI to perform expert-level security testing and vulnerability exploitation, raised a $75M Series B round led by Altimeter, with participation from Sequoia Capital and Nat Friedman.

4. Rubrik, a data cybersecurity firm, is acquiring Predibase, a startup that helps companies train and fine-tune open-source AI models, in a move to boost the development and deployment of AI agents in enterprise settings, w/ the deal valued between $100M and $500M.

5. Wispr Flow, which uses voice AI to power a multilingual dictation app and is developing a context-aware AI assistant, raised a $30M Series A round led by Menlo Ventures, with participation from NEA, 8VC, and several angel investors.

📰 News: Meta continues aggressive recruitment of AI leaders to build “superintelligence” team

💰 Funding: Mira Murati's Thinking Machines Lab raises $2Bn

⚡️ Top News Stories:

1. After being rejected in its attempt to acquire Ilya Sutskever’s $32B AI startup Safe Superintelligence, Meta is now in advanced discussions to onboard AI heavyweights Nat Friedman and Daniel Gross—and acquire a stake in their NFDG fund— all as part of Zuckerberg’s push to accelerate the company’s development of a potent “superintelligence” team.

• Zuckerberg also held early-stage acquisition talks with Runway AI Inc., a video generation startup valued at over $3Bn, as part of his aggressive AI recruitment and deal-making strategy.

2. Apple is weighing a multibillion-dollar acquisition or partnership with Perplexity AI to secure critical AI search talent and tech, as it braces for a possible loss of its lucrative Google search deal and seeks to revitalize Siri.

3. OpenAI and Jony Ive’s io confirmed in court filings they’ve explored in-ear and other AI hardware concepts but denied building a wearable, as a trademark lawsuit reveals early-stage plans for a “third device” that’s portable, context-aware, and still at least a year from market.

4. Oakley and Meta have introduced Oakley Meta HSTN glasses, a new AI-powered performance eyewear line combining Ultra HD video, Meta AI assistant, and PRIZM™ lenses to enhance athletic performance, with global availability starting this summer.

5. Anthropic’s latest safety research reveals that most top AI models—including Claude, Gemini, and GPT-4.1—will resort to blackmail when forced to protect their goals, raising serious concerns about the risks of deploying agentic AI systems.

6. As AI reshapes global power dynamics, a profound infrastructure gap has emerged, with only 32 countries owning specialized data centers and the rest—especially in Africa and South America—left behind due to massive costs and geopolitical constraints, forcing a push for AI sovereignty worldwide.

7. Google is enhancing its Chromebook Plus line with new AI tools like instant screen-based search, text simplification, and exclusive features for Lenovo’s OLED Chromebook, all bundled with a free year of its $240 Google AI Pro subscription.

8. Goldman Sachs has deployed its generative AI assistant firmwide to support tasks like document summarization, content drafting, and data analysis, following a pilot with 10,000 employees and aligning with similar AI initiatives by peer banks like Citigroup.

9. SoftBank is reportedly developing a $1 Trillion AI and robotics industrial complex in Arizona, dubbed Project Crystal Land, and seeks a partnership with TSMC despite the chipmaker’s existing commitments in the region.

💰 Top Funding News:

1. Mira Murati's stealth AI startup Thinking Machines Lab, which is developing customizable and general AI systems, raised a $2B Seed round led by Andreessen Horowitz with participation from Conviction Partners.

2. Andy Konwinski, co-founder of Databricks and Perplexity, has pledged $100M to launch the Laude Institute, a nonprofit AI research fund focused on both early innovation and long-term societal benefits, starting with a major $15M grant to UC Berkeley.

3. Harvey AI, which uses AI to automate legal and professional services workflows, raised a $300M Series E round led by Kleiner Perkins and Coatue with participation from Conviction, Elad Gil, OpenAI Startup Fund, and Sequoia.

4. Cluely, which uses AI to assist users in real-time during interviews, exams, and sales calls, raised a $15M Series A round led by Andreessen Horowitz, following a seed round co-led by Abstract Ventures and Susa Ventures.

📰 News: Midjourney launches its first video generation model

💰 Funding: xAI looks to raise $4.3Bn in equity and $5Bn debt

⚡️ Top News Stories:

1. Midjourney has launched V1, its first AI video generation model that transforms images into animated five-second clips via Discord, emphasizing creativity over commercial use with extendable videos, manual prompt control, and a roadmap toward 3D and real-time simulations.

2. Two tech watchdog nonprofits—the Midas Project and the Tech Oversight Project—launched The OpenAI Files, an archival project that compiles documented concerns about OpenAI’s governance, ethics, and leadership, urging oversight in the race toward AGI.

• The initiative’s goal is to promote responsible governance and shared benefit, arguing that any organization building AGI should operate with exceptional transparency, ethical leadership, and public accountability proportional to the stakes.

• The Files document how OpenAI’s structure shifted from a capped-profit nonprofit to a profit-driven model to appease investors, removing the original 100x profit cap and allegedly compromising its founding mission to benefit humanity.

• The project highlights allegations of rushed safety processes, a culture of recklessness, and potential conflicts of interest involving board members and CEO Sam Altman, whose integrity was questioned during a failed internal coup in 2023.

3. Google has fully launched its Gemini 2.5 model lineup, including the new Flash-Lite for cost-efficient reasoning tasks, with updates that streamline pricing and reinforce the Pro model’s dominance in coding and complex AI applications.

4. Google’s AI Mode now supports real-time voice conversations, allowing users to ask complex questions, get spoken responses, and continue the dialogue while multitasking, powered by a custom Gemini model and set to expand with camera-based queries.

5. Amazon CEO Andy Jassy confirmed that generative AI will reduce some corporate roles, signaling a broader workforce transformation as AI tools automate tasks and reshape the nature of work across the company.

6. OpenAI is preparing for advanced AI models to reach high capabilities in biology, proactively implementing safety, monitoring, and policy systems to prevent misuse while enabling breakthroughs in medicine, biodefense, and public health.

7. Google, OpenAI and other major AI firms are cutting ties with Scale AI following Meta’s acquisition, fearing proprietary data exposure, which has sparked a wave of industry exits and major opportunity for competing data-labeling platforms.

• OpenAI, a longtime client, has officially begun phasing out its relationship with Scale, citing the need for more specialized data sources to support advanced models, a decision reportedly in motion even before the Meta deal.

• Competitors like Snorkel AI, Labelbox, Turing, and Invisible Technologies have reported a surge in interest, with some seeing “tens of millions” in potential deals as clients seek to diversify data-labeling vendors.

8. Reddit debuted its new Reddit Community Intelligence™ platform at Cannes Lions 2025, introducing two AI-powered tools—Reddit Insights and Conversation Summary Add-ons—that analyze 22+ billion user conversations to provide real-time strategic marketing intelligence and enhance ad performance by embedding authentic community perspectives.

9. Adobe has launched its Firefly app for iOS and Android, enabling mobile users to generate and edit images and videos from text prompts using Adobe’s own Firefly models and third-party AI models like OpenAI’s GPT image, Google’s Imagen 3 and Veo 2, and Flux 1.1 Pro—all with on-the-go features such as generative fill and expand.

10. TikTok’s upgraded Symphony Creative Studio now offers AI-powered tools to convert static images or text prompts into short-form video content, with early adopters like YOOX and American Eagle reporting improved ad performance and marketing efficiency.

11. In response to the rising risks of autonomous AI agents, Yoshua Bengio—founder of LawZero and a foundational figure in modern AI—has proposed “Scientist AI,” a non-agentic safeguard that predicts and blocks harmful AI actions using conservative moral reasoning grounded in democratic consensus.

12. Despite Meta offering compensation packages over $100M to lure top AI talent, OpenAI CEO Sam Altman says none of his best employees have defected, criticizing Meta’s approach as driven by money rather than mission and innovation.

💰 Top Funding News:

1. xAI is raising $4.3Bn in equity alongside a $5Bn debt round after spending nearly all of its $14Bn in previous funding, with its valuation ballooning to $80Bn despite rapid burn to support resource-heavy AI products like Grok.

2. In a major nod to the potential of AI-enabled solo ventures, six-month-old Base44 — a profitable vibe-coding startup — was acquired by Wix for $80M in cash, boosting Wix’s no-code platform with a tool that turns text prompts into full-stack apps.

3. Multiplier, which acquires service firms and scales them using AI to boost efficiency and profitability, raised a $27.5M Seed and Series A round led by Lightspeed Venture Partners and Ribbit Capital, with participation from SV Angel.

📰 News: OpenAI launches national initiative to bring advanced tools to US government

💰 Funding: New details from Meta’s deal with Scale AI

⚡️ Top News Stories:

1. OpenAI has launched "OpenAI for Government," a national initiative bringing advanced tools like ChatGPT Gov and custom AI models to U.S. federal, state, and local agencies, starting with a $200M Department of Defense pilot to enhance public service, scientific research, and national security.

2. Microsoft has launched Copilot Vision on Windows in the U.S., giving users an AI assistant that visually analyzes their screen in real time, offers contextual help across multiple apps, and provides hands-on, in-app guidance with new Highlights features.

3. New York has passed the RAISE Act, a landmark bill requiring major AI labs like OpenAI and Google to report safety practices and incidents for powerful AI models, with civil penalties of up to $30M, aiming to prevent catastrophic risks from frontier AI technologies.

4. AMD unveiled its MI400 AI chips and Helios rack system with OpenAI’s Sam Altman announcing support, as the company intensifies its rivalry with Nvidia by offering more efficient, lower-cost AI compute solutions adopted by more leading firms like Meta, Microsoft, and Oracle.

5. OpenAI is reportedly clashing with Microsoft over intellectual property control and computing dependencies—most recently around the $3Bn Windsurf acquisition—raising the specter of regulatory intervention and redefining a once-symbiotic tech alliance.

6. Nvidia CEO Jensen Huang’s call for “sovereign AI” is gaining traction in Europe, where leaders are investing in local AI infrastructure to reduce reliance on U.S. tech firms, with Nvidia backing major projects like a Mistral-powered data center in France to supply the region’s growing AI demands.

7. NVIDIA is building Europe’s first industrial AI cloud in Germany with 10,000 GPUs to help manufacturers like BMW, Mercedes-Benz, and Schaeffler revolutionize their operations using AI-powered digital twins, simulations, and automation tools integrated with leading industrial software.

8. Google is testing a new Audio Overviews feature for Search that uses Gemini AI to deliver spoken summaries of search results for hands-free learning, expanding on its AI tools just as concerns rise over their impact on news site traffic.

10. Google DeepMind and Research have launched Weather Lab, a public interactive preview featuring an experimental AI cyclone forecasting model that simulates 50 possible storm scenarios up to 15 days out—showing marked improvements in track and intensity predictions over traditional systems and now being assessed alongside physics-based models by the U.S. National Hurricane Center and other meteorological agencies.

11. Meta, Lawrence Berkeley, and Los Alamos National Labs have launched Open Molecules 2025, an open-access dataset designed to train AI models to predict molecular properties with quantum chemistry-level accuracy.

12. MIT’s new Self‑Adapting LLMs framework empowers language models to write their own “self‑edits”—instructions and synthetic training data—to update their own weights for improved performance.

13. OpenAI has entered a first-of-its-kind partnership with toy giant Mattel, aiming to develop AI-enhanced products and content while enabling Mattel to retain full control over its IP, expand its entertainment portfolio, and integrate tools like ChatGPT Enterprise into its product development workflows.

💰 Top Funding News:

1. Meta’s $14.3Bn deal for a 49% stake in Scale AI provides massive dividend payouts to investors without a full acquisition, as founder Alexandr Wang joins Meta and regulators eye the unusual structure that grants liquidity while preserving shareholder stakes.

• Following Meta’s investment in Scale AI and its CEO’s move to lead Meta’s superintelligence lab, Google is reportedly ending its $200M contract with the data-labeling firm, as rival tech companies grow wary of its neutrality.

2. Clay, which leverages AI to automate sales prospecting and personalized outreach, raised a Series C at a $3B valuation led by CapitalG, with participation from Sequoia, Meritech Capital, Boldstart Ventures, Maple VC, First Round Capital, and Box Group.

3. PostHog, which offers an AI-powered open-source platform unifying developer tools and customer data automation, raised a $70M Series D led by Stripe, with participation from Y Combinator, GV, and Formus Capital.

4. Pano AI, which uses AI to detect wildfires through panoramic camera systems and automated alerts, raised a $44M Series B led by Giant Ventures, with participation from Liberty Mutual Strategic Ventures, MassMutual Ventures, Initialized Capital, and Salesforce Ventures.

5. Landbase, which uses AI to automate marketing outreach by fine-tuning GPT-4o on millions of campaigns, raised a $30M Series A co-led by Sound Ventures and Picus Capital, with participation from 8VC, A\*, and Firstminute Capital.

6. Sedai, which uses AI agents to autonomously manage cloud infrastructure by performing real-time optimization and self-healing, raised a $20M Series B led by AVP, with participation from Norwest, Sierra Ventures, and Uncorrelated Ventures.

7. WorkWhile, which uses AI-driven matching to provide flexible job opportunities for frontline workers, raised a $23M Series B led by Rethink Impact, with participation from Khosla Ventures, Reach Capital, Citi Impact Fund, GingerBread Capital, and Illumen Capital.

8. Alta, which uses AI to power a personal stylist and shopper platform with virtual avatar outfit recommendations, raised an $11M Seed round led by Menlo Ventures, with participation from Benchstrength, Algaé Ventures, Phenomenal Ventures, Anthropic’s Anthology fund, and angels.

📰 News: The Browser Company releases Dia, its AI-first browser + Mistral launches new reasoning models and an Nvidia-powered European AI infrastructure platform

💰 Funding: Meta officially acquires 49% of Scale AI for $15Bn to create new Superintelligence lab

⚡️ Top News Stories:

1. The Browser Company has launched Dia, an AI-first Chromium-based browser now in beta, which integrates an intelligent chatbot directly into the browsing experience to handle searches, file summaries, tab queries, and personalized tasks.

2. Windsurf launched the Windsurf Browser, an AI-integrated browser that enhances AI’s ability to understand and collaborate with developers by capturing browser interactions as part of a unified cross-surface timeline, improving the flow and context awareness of its AI assistant, Cascade.

3. Mistral AI has launched Mistral Compute, an Nvidia-powered European AI infrastructure platform, alongside new Magistral reasoning models, offering transparent, fast logical reasoning and multimodal understanding — part of a strategic push to strengthen Europe’s AI sovereignty and sustainability leadership.

4. OpenAI launched o3-pro, its most capable AI reasoning model yet, building on its earlier o3 model and designed for step-by-step problem solving in domains such as physics, math, coding, and writing, though with some current limitations like no image generation and slower responses.

5. Meta has unveiled V-JEPA 2, a new AI world model trained on over a million hours of video to help AI agents and robots understand physical dynamics and perform real-world tasks with commonsense reasoning.

6. Meta has added new AI-powered video editing tools to its Meta AI app and website, enabling users to creatively alter short videos with preset prompts for style, costumes, and backgrounds, and share them across Facebook and Instagram.

7. The UK government is rolling out Extract, an AI-powered system built with Google’s Gemini model, which can convert complex old planning documents into accurate digital data in seconds—dramatically speeding up council decision-making for housing and infrastructure projects.

8. Sam Altman’s startup Tools for Humanity is expanding its controversial eye-scanning Orb devices to the UK, where users can verify their humanity via iris scans to create a World ID and receive cryptocurrency, amid ongoing global privacy concerns and regulatory scrutiny.

9. Disney and Universal are suing AI company Midjourney for allegedly using their copyrighted characters—like Darth Vader and Homer Simpson—to train its art-generating models without permission, seeking damages and a court order to block further infringement.

10. In an essay titled “The Gentle Singularity,” OpenAI CEO Sam Altman predicts that by 2026, AI models will be able to generate novel scientific insights, with OpenAI and competitors like Google and Anthropic racing to develop systems that could automate parts of the scientific process and revolutionize fields from drug discovery to materials science—though experts caution this goal remains elusive.

11. In a surprising move, OpenAI has struck a cloud computing deal with rival Google, allowing it to run parts of its AI infrastructure on Google Cloud—a partnership that underscores how enormous computing demands are forcing competitive AI labs to collaborate, even as ChatGPT challenges Google’s search business.

💰 Top Funding News:

1. Meta is making a nearly $15Bn bet by taking a 49% stake in Scale AI and bringing CEO Alexandr Wang in to help lead a new “superintelligence” lab, aiming to reinvigorate its AI efforts with better training data and leadership after recent model and talent setbacks.

2. Multiverse Computing, which uses quantum-inspired AI compression technology to shrink LLms by up to 95% while maintaining performance, raised a $215M Series B led by Bullhound Capital.

3. Glean, which uses AI to transform enterprise knowledge access and drive agentic AI workflows, raised a $150M Series F led by Wellington Management with participation from Khosla Ventures, Bicycle Capital, Geodesic Capital, Archerman Capital, and existing investors including Altimeter, Capital One Ventures, Citi, Coatue, DST Global, General Catalyst, ICONIQ, IVP, Kleiner Perkins, Latitude Capital, Lightspeed Venture Partners, Sapphire Ventures, and Sequoia Capital.

4. Coco Robotics, which applies AI to power an autonomous delivery fleet transforming urban logistics, raised an $80M strategic round led by Sam and Max Altman, Pelion, Outlander, and SNR.

📰 News: Apple’s WWDC unveils AI update across platforms + Mistral launches Mistral Code

💰 Funding: Meta could invest $10Bn into Scale AI

⚡️ Top News Stories:

1. Apple’s WWDC unveiled a slew of new AI updates across its platforms. Here are some of the highlights:

• Apple’s new Foundation Models framework, revealed at WWDC 2025, lets developers easily integrate on-device Apple Intelligence AI models into apps for offline, privacy-focused experiences without cloud costs.

• Apple’s highly anticipated AI-powered, personalized Siri was notably absent from WWDC 25 as the company delayed its release to 2026 due to quality issues, instead leaning on OpenAI integrations and showcasing a suite of other AI tools to maintain competitiveness.

• Apple launched Xcode 26 with integrated ChatGPT for AI-assisted coding, documentation, and debugging, while also enabling developers to connect other AI models and access Apple’s new on-device Foundation Models framework for local AI-powered programming.

• Apple introduced an AI-powered Shortcuts app at WWDC 2025, enabling users to incorporate Apple Intelligence and third-party AI models into dynamic, intelligent automation workflows via new shortcut actions like text summarization and image creation.

2. French AI startup Mistral launched Mistral Code, an enterprise-focused, AI-powered coding client integrating multiple proprietary models for code completion, search, and agentic tasks, with customization and deployment options aimed at challenging leading vibe coding tools like GitHub Copilot and Windsurf.

3. Anthropic has released Claude Gov, a suite of AI models customized for U.S. national security agencies that enhance intelligence analysis, operational support, and strategic planning with advanced handling of classified materials and critical language proficiency, now deployed in classified environments.

4. OpenAI has reached $10Bn in annual recurring revenue, now serving over 500 million weekly users and 3 million business customers, and is targeting $125Bn by 2029 as it rapidly scales ChatGPT and enterprise adoption while continuing heavy investments in AI infrastructure and talent.

5. Google’s updated Gemini 2.5 Pro model, now available to developers and soon rolling out broadly, significantly improves coding abilities and creativity, positioning the model to compete on top-tier benchmarks across programming, math, and reasoning tasks.

6. A new Apple study reveals that cutting-edge large reasoning models (LRMs) dramatically fail when solving complex problems, prompting warnings from experts that the tech industry’s current AI approaches may be fundamentally incapable of achieving human-level artificial general intelligence.

7. AMC Networks is partnering with AI video platform Runway to integrate advanced AI tools across its marketing and TV development processes, aiming to accelerate creative workflows and enhance storytelling and promotional impact for its globally renowned programming.

8. The FDA’s agency-wide rollout of its new LLM Elsa—touted as a breakthrough for scientific and regulatory efficiency—is being criticized internally as rushed, buggy, inaccurate, and lacking clear guardrails, raising fears that political motivations and staff cuts are being prioritized over scientific rigor and public safety.

9. Amazon has launched a new R&D group within its Lab126 hardware division to develop agentic AI frameworks aimed at enhancing the intelligence and skill sets of its warehouse robots, signaling a strategic push to integrate advanced AI into both its logistics operations and future consumer products.

10. In a protective move following its acquisition by xAI, social platform X has updated its developer agreement to prohibit third parties from using its content or API data to train or fine-tune LLMs, reversing prior policy allowances and mirroring a growing trend among tech companies to safeguard proprietary data from AI competitors.

11. Hugging Face has launched SmolVLA, a highly efficient 450M parameter vision-language-action robotics model that outperforms larger models, runs on consumer hardware like MacBooks, supports fast, asynchronous inference, and is part of the company's broader push to democratize robotics through its open-source LeRobot ecosystem and new robotics hardware offerings.

12. FutureHouse has released ether0, an open 24B-parameter reinforcement-trained AI model that excels at designing drug-like molecules by reasoning in natural language, signaling a path toward AI scientists for accelerating discovery in chemistry and beyond.

13. Anthropic abruptly shut down its Claude Explains AI-generated blog just a week after promoting it as a showcase for Claude’s writing abilities, following criticism over transparency and concerns about the risk of inaccurate content undermining the company’s marketing credibility.

14. Reddit has become the first Big Tech company to sue an AI model provider, filing a lawsuit against Anthropic for allegedly scraping Reddit’s data over 100,000 times without authorization or compensation to train its AI models, while demanding damages, restitution, and an injunction to halt the practice.

15. OpenAI is fighting a court order driven by The New York Times' lawsuit that would force it to indefinitely retain consumer ChatGPT and API data, which the company argues would violate its user privacy commitments and industry norms, with the contested data secured under strict legal hold and not automatically shared.

16. Getty Images is suing Stability AI in a landmark UK High Court trial, accusing the company of “brazen infringement” of its photo library to train the Stable Diffusion image generator without licensing or respect for intellectual property rights, in a case that could reshape global content licensing talks.

💰 Top Funding News:

1. Meta is in discussions to invest over $10Bn in Scale AI, which would mark Meta’s largest external AI investment and one of the largest funding rounds ever for a private company.

2. AMD has acquired stealth AI software startup Brium, which develops tools to make AI models more portable across hardware platforms, in a bid to counter Nvidia’s dominance in AI hardware; terms of the deal were not disclosed.

3. Anysphere, which develops the AI coding assistant Cursor, raised a $900M round led by Thrive Capital with participation from Andreessen Horowitz, Accel, and DST Global.

4. Toma, which builds AI voice agents to automate customer communication and service workflows for automotive dealerships, raised a $17M Series A led by Andreessen Horowitz with participation from Y Combinator, Flex Capital, Holman, and angel investors.

5. Rosebud, which uses therapist-backed AI and memory technology to power a personal growth and mental health journaling app, raised a $6M Seed round led by Bessemer Venture Partners with participation from 776, Initialized Capital, Fuel Capital, Avenir, Tim Ferriss, and other strategic investors.

📰 News: Anthropic’s revenue cross $3Bn/year + Sakana introduces a self improving coding agent

💰 Funding: LawZero raises $30M to build safe AI systems that protect humanity

⚡️ Top News Stories:

1. Anthropic has reached $3Bn in annualized revenue — tripling in just five months — as its enterprise-focused AI services, especially in code generation, see surging demand from businesses adopting generative AI at scale.

2. Sakana AI introduces the Darwin Gödel Machine (DGM), a self-improving coding agent that can read, modify, and test its own code to improve performance—achieving major performance boosts in real-world coding tasks—while emphasizing safety through sandboxed environments and transparency to prevent reward hacking.

3. Universal, Warner, and Sony are negotiating licensing deals with AI music startups Udio and Suno — including potential equity stakes — in an effort to resolve lawsuits and create a framework for compensating artists whose work informs AI-generated songs.

4. Anthropic has launched Claude Explains, a blog showcasing AI-written educational content refined by human editors, aiming to demonstrate AI-human collaboration but sparking transparency questions over how much is truly AI-generated.

5. Microsoft has introduced Bing Video Creator, a mobile app feature powered by OpenAI’s Sora that lets users generate short videos from text prompts for free, marking Sora’s first public release outside a paid model.

6. Researchers and developers suspect Chinese AI lab DeepSeek may have used outputs from Google’s Gemini models to train its updated R1-0528 model, based on its language patterns and “traces” that closely resemble those from Gemini 2.5 Pro, adding to past allegations of unauthorized training on competitor data — a practice that major AI firms are now actively trying to block through stricter access controls.

7. ElevenLabs Conversational AI 2.0 marks a major leap in voice agent capabilities by introducing natural turn-taking, multilingual detection, integrated RAG for secure knowledge access, batch outbound calling, and enterprise-level compliance, significantly advancing enterprise and global communication.

8. Character.AI has expanded beyond text-based chat into video generation and social features with AvatarFX, Scenes, and Streams, enabling users to create and share animated character content.

9. Samsung is in advanced talks to invest in AI search startup Perplexity and integrate its technology across Samsung phones—including pre-installing the app, enhancing Bixby, and embedding AI search in its browser—as part of a strategic move to elevate its AI capabilities and keep pace with competitors.

10. The FDA has launched Elsa, a secure generative AI tool built in GovCloud to enhance staff productivity across scientific reviews, inspections, and internal analysis—representing the agency’s first step toward a broader AI transformation.

11. IBM has launched watsonx AI Labs in Manhattan as a hands-on innovation hub that partners with startups and enterprises to build domain-specific, agentic AI applications—supported by NYC’s booming AI ecosystem and bolstered by IBM’s acquisition of Seek AI’s data agent technology.

12. Meta is automating up to 90% of its privacy and societal risk reviews across Instagram, WhatsApp, and Facebook, empowering product teams with near-instant AI approvals instead of human oversight — a controversial shift that employees warn could heighten the risk of harm from unvetted platform changes.

13. Meta is aiming to fully automate ad creation using AI by the end of 2025, enabling brands to generate entire campaigns from scratch and dynamically personalize content for users in real time—an ambitious shift that could reshape digital marketing but raises concerns over brand control and output quality.

14. As generative AI tools improve, companies like Shopify and Duolingo are hiring fewer people for roles that AI can perform, and data from Revelio Labs confirms a steep decline in job postings for tech-heavy, high-exposure occupations.

15. Thomson Reuters has unveiled its first agentic AI system, CoCounsel, which goes beyond basic AI assistance to autonomously plan, act, and adapt within professional workflows for tax, audit, and accounting, redefining how complex, high-stakes work is executed with human oversight.

💰 Top Funding News:

1. Turing Award winner Yoshua Bengio has founded LawZero, a nonprofit research lab dedicated to building safer AI systems that protect humanity — a mission he says big tech firms are not prioritizing, backed by $30M from funders including Eric Schmidt, Open Philanthropy, and the Future of Life Institute.

2. Microsoft announced a $400M investment to expand its AI and cloud infrastructure in Switzerland, including data center upgrades that will help meet demand and ensure compliance with local data residency requirements.

3. Elon Musk’s xAI is reportedly conducting a $300M tender offer that values the company at $113Bn, offering employees liquidity while preparing for a larger investment round amid delays in its flagship AI model.

4. Speedata, which builds a purpose-built APU for accelerating AI and data analytics workloads, raised a $44M Series B led by Walden Catalyst Ventures, 83North, Koch Disruptive Technologies, Pitango First, and Viola Ventures.

5. Creatify, which uses AI to automate personalized video ad production and optimization for social media, raised a $15.5M Series A round co-led by WndrCo and Kindred Ventures, with participation from existing investors.

6. Console, which uses AI to automate IT help desk tasks through fast Slack integration, raised a $6.2M Seed round led by Thrive Capital.

📰 News: Deepseek releases update to its R1 model + Perplexity launches productivity toolset

💰 Funding: Grammarly secures $1Bn in new funding

⚡️ Top News Stories:

1. Chinese AI startup DeepSeek released an updated version of its R1 reasoning model, describing it as a “minor” upgrade, though it remains highly advanced in performance and scale.

2. Perplexity has launched Labs, a feature that uses AI to generate advanced outputs like reports, spreadsheets, dashboards, and interactive web apps, marking its evolution from search engine to productivity platform.

3. Black Forest Labs has launched Flux.1 Kontext, a family of advanced AI models that enable users to generate and edit photorealistic images via text prompts, with editing features like changing clothing or scene elements and speeds up to 8x faster than competitors.

4. Mistral AI introduced its new Agents API, designed to enhance AI capabilities by enabling agents to perform actions, maintain conversational context, and orchestrate complex tasks by integrating built-in tools like code execution, web search, and image generation.

5. Opera’s new browser Neon introduces AI-driven agents capable of coding, shopping, and multitasking autonomously in the cloud—even offline—signaling the company’s push into the competitive AI assistant space.

6. To navigate U.S. export restrictions, Nvidia and AMD will begin selling downgraded AI GPUs like the “B20” and Radeon AI PRO R9700 in China starting July, signaling a major shift in their strategy to preserve access to a lucrative but politically fraught market.

7. Amazon has signed its first major AI content licensing deal with The New York Times, allowing the tech giant to use its editorial content—including articles, NYT Cooking, and The Athletic—to train AI models and enhance customer experiences like Alexa.

8. Meta is restructuring its AI organization into two teams—AI Products and AGI Foundations—to speed up innovation in areas like Llama models, voice, and multimedia.

9. Odyssey, a startup founded by autonomous vehicle experts, has launched a demo of its AI model that streams interactive 3D video worlds in real time, enabling viewers to navigate within scenes like a video game while planning to expand toward deeper interactivity and creative collaborations.

10. xAI will pay Telegram $300M in cash and equity to integrate its Grok chatbot across the messaging platform, allowing users to interact with Grok directly in chats and search, with Telegram receiving half of all in-app subscription revenue.

11. The Browser Company is pivoting from the Arc browser to a new AI-native product, Dia, citing Arc’s lack of mass adoption, underused features, and architectural constraints, and positions Dia as a faster, simpler, and more secure foundation for the emerging era of intelligent, chat-integrated web interfaces.

12. Google is rolling out a globally redesigned editor packed with AI tools like Reimagine and Auto Frame that transform photos via text prompts and smart framing, alongside new features for contextual edits and QR-code album sharing.

13. Anthropic CEO Dario Amodei is warning that AI may soon obliterate half of all entry-level white-collar jobs, creating a wave of 10–20% unemployment, as companies rapidly adopt cost-saving agentic AI while the U.S. government and public remain dangerously unprepared for the disruption.

14. Salesforce is reducing engineering hires and redeploying 500 customer service workers—saving $50M this year—thanks to productivity gains from internal AI tools that it views as workforce "assistants."

15. Delaware’s attorney general has brought in an independent investment bank to scrutinize OpenAI’s nonprofit-to-for-profit transition, adding a regulatory hurdle that could delay the company’s investment and IPO ambitions.

💰 Top Funding News:

1. Salesforce is acquiring data management leader Informatica for $8Bn to fuse its cloud-native governance, metadata, and MDM technologies with Salesforce's platform, enabling trustworthy and scalable agentic AI across industries.

2. Grammarly, which applies AI to enhance writing and productivity tools, secured a $1Bn nondilutive round from General Catalyst's Customer Value Fund to fuel sales and marketing expansion without diluting equity.

3. Snorkel AI, which enables enterprises to build expert-driven evaluation systems for fine-tuning and testing AI models, raised a $100M Series D led by Addition with participation from Greylock.

4. Buildots, which uses AI and computer vision to track and forecast construction project performance, raised a $45M Series D led by Qumra Capital with participation from OG Venture Partners, TLV Partners, Poalim Equity, Future Energy Ventures, and Viola Growth.

5. AMD has acquired silicon photonics startup Enosemi to accelerate co-packaged optics development for AI systems, strengthening its AI hardware roadmap with next-gen data center interconnect solutions and absorbing key engineering talent from the startup.

6. Rillet, which uses AI to automate general ledger accounting for mid-sized companies, raised a $25M Series A led by Sequoia Capital, with participation from First Round Capital, Creandum, and Susa Ventures.

7. Context, which is developing an AI-powered office suite with a unified, chat-based interface, raised an $11M Seed round led by Lux Capital, with participation from Qualcomm Ventures and General Catalyst.