|
OpenAI launches GPT-5.6 family of models with government greenlight
OpenAI has received the Trump administration's approval to publicly release GPT-5.6, which CEO Sam Altman called 'the best model we have ever produced.' The new model family promises improvements across a range of areas including cybersecurity, and will continue to power Microsoft Copilot 365 as the preferred model amid partnership speculation. The release comes alongside OpenAI's broader 'ChatGPT Work' announcement targeting enterprise users.
|
|
Meta launches Muse Spark 1.1 model and API to compete in AI coding
Meta has released Muse Spark 1.1, a significant upgrade from its first in-house model, now available to developers via the Meta Model API for integration into AI coding software. The model is pitched for handling large agentic workloads, bug fixing, and large-scale code migrations that enterprises increasingly demand. Meta claims the new version is a 'step-change' improvement over the original Muse Spark.
|
|
Fidji Simo steps down from OpenAI's No. 2 role due to illness
OpenAI's Fidji Simo is departing her full-time role leading the company's AGI work and transitioning to a part-time advisory position after her medical leave for a neuroimmune condition proved longer than expected. The leadership vacuum comes at a critical time as OpenAI eyes a potential IPO and competes with Anthropic in the enterprise market. Simo had only recently taken on the AGI chief role before needing to step back.
|
|
Anthropic, OpenAI, and SpaceX IPOs set to dwarf 25 years of U.S. tech exits
Three major AI-related IPOs from Anthropic, OpenAI, and SpaceX are projected to generate more value than all U.S. venture-backed tech exits combined since the year 2000. The milestone underscores the unprecedented scale of capital flowing into frontier AI companies. The figures come as all three companies are reportedly preparing for public market debuts.
|
|
Elon Musk promises not to cut off Anthropic access on Mythos/Fable infrastructure
Elon Musk has publicly praised the Mythos/Fable AI infrastructure platform and pledged that Anthropic would not be denied access to host its models there. With approximately $40 billion in potential revenue at stake, Musk's assurances are aimed at convincing Anthropic to trust his platform. The statement raises questions about competitive dynamics between Musk's ventures and rival AI labs.
|
|
Lyzr AI agent startup raises $100M after using its own agent to run the fundraise
Enterprise AI agent startup Lyzr has closed a $100 million funding round, notable for having used its own AI agent to manage the fundraising process. The stunt served as a live demonstration of the company's core product capabilities for potential investors. Lyzr builds AI agents designed to automate complex enterprise workflows.
|
|
Paris AI voice startup Gradium raises $100M seed round backed by Nvidia
Gradium, a Paris-based AI voice startup, has raised a $100 million seed round with backing from Nvidia. The company plans to use the funding to open a Bay Area office and compete for AI talent in the world's leading AI ecosystem. The raise is one of the largest seed rounds in AI voice technology.
|
|
Nvidia caught in the competitive compute marketplace it helped create
An analysis argues that Nvidia's success in proving the value of AI compute has attracted a wave of competitors and adjacent players eager to profit from the infrastructure boom it pioneered. The company now faces pressure from cloud providers building custom silicon and from a market increasingly focused on cost efficiency. Simpler technologies and less prominent companies are reportedly reaping significant rewards on the margins of the GPU economy Nvidia built.
|
|
Can AI deliver on its $3 trillion economic promise?
The ongoing debate over AI's return on investment has intensified, with even larger figures now being cited and greater consequences attached to the outcome. Analysts and industry observers are questioning whether the massive capital expenditures on AI infrastructure will translate into measurable productivity and revenue gains. The stakes of the ROI question are higher than ever as enterprise AI spending continues to accelerate.
|
|
Nandan Nilekani steps back from GP role at Fundamentum as firm launches $200M third fund
Infosys co-founder Nandan Nilekani is stepping down from his general partner role at Indian VC firm Fundamentum as it launches a $200 million third fund targeting AI and fintech startups. Nilekani will remain as the firm's anchor investor while Fundamentum expands its leadership team. The fund will focus on technology startups in India.
|
|
OpenAI shuts down ChatGPT Atlas browser, pivots agentic features to desktop and Chrome
OpenAI is sunsetting Atlas, its AI-powered browser capable of performing tasks on users' behalf, less than a year after its October launch. Agentic browsing features are being moved to OpenAI's desktop app and a new Chrome extension rather than abandoned entirely. The shutdown was announced as part of the company's broader ChatGPT Work rollout.
|
|
Anthropic launches Claude 'Reflect' feature showing users their AI usage data
Anthropic has introduced a 'Reflect' feature for its Claude chatbot, giving users a monthly breakdown analysis of their usage data in the style of Spotify Wrapped. The dashboard visualizes how users interact with the AI, while critics note it subtly reinforces reliance on Anthropic's product. The feature is rolling out to Claude users starting Thursday.
|
|
Ollama raises $65M as open-source local AI developer tool nears 9 million users
Ollama, the popular open-source tool that lets developers run AI models locally on their PCs, has raised $65 million in funding backed by Benchmark. The platform has accumulated nearly 9 million users, 176,000 GitHub stars, and approximately 17,000 forks. The raise signals strong investor appetite for developer-focused, locally-run AI infrastructure.
|
|
FL Studio 2026 upgrades AI chatbot Gopher into a full assistant engineer
The new FL Studio 2026 release significantly expands Gopher, the music production software's AI chatbot, from a basic instruction manual into an interactive assistant that can perform engineering tasks within the DAW. The upgraded Gopher can now take actions and make changes to a user's project rather than simply providing text instructions. The update positions AI as a hands-on collaborator in music production workflows.
|
|
New York Times accuses OpenAI of hiding evidence in ChatGPT copyright trial
The New York Times and other news publishers have filed a motion for sanctions against OpenAI, alleging the company concealed tools and datasets that could identify copyrighted journalism in ChatGPT outputs. The accusation escalates an already high-profile copyright lawsuit between major media organizations and OpenAI. Publishers argue the hidden evidence is directly relevant to proving their claims.
|
|
Microsoft's carbon emissions rose 25% in 2025 due to AI expansion
Microsoft's 2026 sustainability report reveals the company's carbon emissions increased 25 percent in 2025, reaching 34 million metric tons without select interventions. The rise is attributed primarily to the rapid expansion of its data center infrastructure to support AI workloads. The increase puts Microsoft further behind its own stated climate goals.
|
|
Google to label ads created or edited with AI across Search, Discover, and YouTube
Google is introducing a 'created or edited with AI' label for ads appearing on Google Search, Google Discover, and YouTube, visible through the 'My Ad Center' settings. The disclosure requirement expands beyond election ads, which previously were the only category required to declare AI-generated content. The update was announced Thursday and marks a step toward greater transparency in AI-generated advertising.
|
|
Microsoft using AI to find more security vulnerabilities, expanding Patch Tuesday updates
Microsoft announced it is now using AI to identify potential security issues earlier in the development cycle, which will result in a higher volume of fixes included in each monthly Patch Tuesday release. The initiative comes as both hackers and security researchers increasingly leverage AI tools to discover and exploit vulnerabilities. Windows 11 users can expect larger but more comprehensive security updates going forward.
|
|
How to stop Meta's AI from using your Instagram photos
Meta's Muse Image tool allows users to generate AI images using photos pulled from public Instagram accounts, raising privacy concerns for the platform's users. Any public profile's images can be used as source material when another user tags that account in an AI generation request. TechCrunch has published guidance on how Instagram users can adjust their settings to limit this use of their images.
|
|
Meta's new custom AI chips set to enter production in September
Meta's next generation of in-house AI chips are scheduled to begin production in September, using a modular design approach to stay adaptable as AI requirements evolve rapidly. The chips are part of Meta's broader strategy to reduce dependence on third-party suppliers like Nvidia. The modular architecture is intended to allow the company to update components as AI workloads change.
|
|
AWS GraphRAG deployment cuts pharmaceutical drug research cycles by 87%
A new AWS GraphRAG deployment has reduced drug research and development cycles in pharmaceutical settings by 87 percent by integrating previously siloed proprietary databases into a unified, queryable knowledge graph. Data gathering and screening phases that previously took over six months per iteration can now be completed far more rapidly. The system also improved the historically low five percent success rate for initial research iterations.
|
|
NHS trials AI blood test to reduce invasive womb cancer checks
Several NHS hospitals are preparing to deploy an AI-powered blood test to help assess women referred for possible womb cancer, potentially reducing the need for invasive diagnostic procedures. Around 90,000 postmenopausal women in England are referred annually by GPs for checks following heavy bleeding, of whom approximately 10,000 are ultimately diagnosed with womb cancer. The AI test could serve as a less invasive preliminary screening step before more intrusive examinations.
|
|
Character.AI launches interactive microdrama series for mobile
Character.AI has announced c.ai Series, a new format of short-form episodic vertical videos that users can watch and interact with on their phones. Unlike traditional microdramas, viewers can chat with characters, ask questions, and roleplay alternative storylines using Character.AI's underlying LLM technology. The move expands the platform beyond chatbots into AI-driven entertainment content.
|