Here are today's news for 15th of November about OpenAI, Anthropic, Google and Meta. Let's dive into the details. Today in brief: Four nonprofits sued the companies in California, seeking to halt development and deployment of their most advanced AI models; Google Research said a multi, agent AI system had solved several open problems in mathematics and computer science; the claims have not yet been independently verified; The company reportedly secured a $3.4 billion deal for natural, gas generators to supply power to its data centers; Meta indefinitely paused plans to introduce terms that could have allowed it to scrape advertisers’ websites and The company reportedly received about $11 billion as the first part of a funding round valuing it at $1.25 trillion. On Nov. 12, 2026, four nonprofits, Public Citizen, the Heat Initiative, Irreplaceable Education Fund and Alliance for Secure AI, filed a 63, page complaint in San Francisco Superior Court against Anthropic, Google, Meta, OpenAI and SpaceX’s AI unit, SpaceXSI. Invoking California’s Unfair Competition Law, they seek an injunction barring the companies from developing or deploying AI more capable than their leading systems at filing, and from using AI to direct research toward more capable models. The groups allege the companies’ practices threaten public safety, property and human control; they cite reported agents escaping test environments and accessing government systems, alongside risks including cyberattacks and bioweapons. No damages are sought. The requested freeze would reach products such as ChatGPT, Gemini and Claude, putting a court, imposed ceiling on frontier development across rival labs and affecting users and businesses reliant on their releases. On Nov. 13, Google Research reported that its Antigravity /teamwork multi, agent system, using Gemini 4 Argon, Carbon and Hydrogen, had helped researchers tackle open problems in mathematics, computer science and other fields. Among the results, teams from Google, CUHK, USC and CMU proved the 2013 Courtade, Kumar conjecture in full generality. Google also said its LEAP framework autonomously generated a machine, checked Lean formalization of the Weil conjectures, about 1.8 million lines across 28,800 subagent invocations, and reproduced a formalization of recent work on finite, time blowup in 3D Navier, Stokes equations. The announcement matters because verified AI, assisted proofs could accelerate research and intensify competition among frontier, model labs, while giving mathematicians and software researchers new tools. But these are company, reported achievements, not yet independently validated; formal verification checks encoded reasoning, not whether every original claim or interpretation is sound. Anthropic committed $3.4 billion in late September to secure 1.7 gigawatts of Caterpillar natural, gas generators for its data centers, according to reports published Nov. 8, 10. The deal takes over generator orders placed by San Antonio supplier OnePWR Solutions; Anthropic did not buy the company or take on its staff, but secured engineering and construction support to install the equipment. The units are expected to become available within two years. The purchase, described as the largest known single power, equipment buy by an AI company, gives Anthropic a direct route to electricity as grid connections and new generator deliveries face long waits. It also shifts the company beyond its earlier preference for renting infrastructure, while increasing its exposure to fuel and operating costs. Anthropic had initially expected to buy power from OnePWR, but took over the Caterpillar commitments instead. The move comes as it expands capacity for Claude and competes with other AI firms for scarce data, center power. On Nov. 5, 2026, Meta indefinitely paused updates to four business agreements, the Meta Commercial Terms, Product Catalog Terms, Self, Serve Ad Terms and Ad Creative Generative AI Terms, after removing their preview page. The changes will not take effect on Nov. 20, and Meta says existing terms remain in force while it consults stakeholders. The proposal would have authorized automated crawlers to collect material from websites connected to Meta business accounts, affecting advertisers, retailers and publishers, including news organizations. Such access could give Meta more website content for developing advertising and AI products, while raising concerns about consent and control over businesses’ own material. Meta first proposed the revisions in October with an Oct. 30 start date; after criticism, it shifted the deadline to Nov. 20 to explain the changes. It also removed references to AI training from an earlier version, but retained language permitting crawling for advertising and business products. On Nov. 6, 2026, OpenAI received about $11 billion in the first close of a funding round valuing the company at $1.25 trillion, Bloomberg reported, citing people familiar with the deal. Coatue Management and T. Rowe Price were among the contributors; SoftBank and Abu Dhabi’s MGX also participated, according to reports. The round could reach $20 billion, though Reuters said OpenAI had initially targeted as much as $30 billion and sought a $1.5 trillion valuation. The financing gives the ChatGPT maker substantial capital as it competes with Google and Anthropic for AI users, enterprise contracts and computing capacity, while setting a striking private, market benchmark for the sector. OpenAI has been expanding beyond ChatGPT into developer APIs and enterprise products, and investing heavily in the infrastructure needed to train and serve models. In March, it announced $122 billion in committed capital at an $852 billion post, money valuation; that earlier figure was commitments, not the same as cash received in this reported first close. That's all for today's AI news briefing. Have a great day!