Claude Fully Formalizes Fermat’s Last Theorem

Anthropic has announced that Claude has completed the first fully computer-verified formalization of Fermat’s Last Theorem, transforming the mathematical proof into a form that can be automatically checked by the Lean proof assistant. Claude worked largely autonomously for 11 days and generated approximately 13 million lines of Lean code, with nearly 29,500 intermediate theorems in the final formalization. The project used dozens of Claude agents collaborating through the Prove2Me platform. Anthropic – Formalizing Fermat's Last Theorem

Why it matters: Claude did not discover a new proof of Fermat’s Last Theorem. The theorem was already proved by Andrew Wiles in the 1990s. The novelty is that AI has managed to transform an extremely complex mathematical proof into a formalization that is fully verified by a computer. Such systems could significantly accelerate proof verification and mathematical research, reducing the risk of subtle errors going unnoticed.

U.S.: The Seattle Times and Newsday Sue OpenAI and Microsoft

The Seattle Times and Newsday have sued OpenAI and Microsoft, accusing the companies of using journalistic articles without permission, including material behind paywalls, in datasets used to develop and operate AI products such as ChatGPT and Copilot. The two publishers argue that the systems can reproduce or paraphrase their content, potentially reducing traffic to their websites and the number of subscriptions. Reuters – The Seattle Times and Newsday sue OpenAI and Microsoft

Why it matters: The lawsuit adds two major publishers to an increasingly broad legal dispute over copyright and AI model training. OpenAI argues that its models are trained on publicly available data and that such use is protected under the fair use doctrine, while publishers are seeking, among other remedies, the destruction of copies, datasets or models containing their copyrighted works. The case could help define the legal boundaries for the use of journalistic content by AI companies.

India to Invest Up to $7.4 Billion in a 1 GW AI Campus

A subsidiary of Tata Consultancy Services (TCS) and its partners intend to invest up to 700 billion rupees, approximately $7.4 billion, to develop an AI data center campus with a capacity of up to 1 GW in Telangana. HyperVault, the TCS subsidiary involved in the project, has secured 264 acres in Hyderabad, and the development will be carried out in multiple phases. The infrastructure is intended for AI companies and hyperscalers and will be capable of supporting GPU systems for training and inference. Reuters – India to invest $7.4 billion in an AI campus

Why it matters: The global AI race is increasingly becoming a race for energy, chips and data centers. A 1 GW campus would place India among the major players in the infrastructure required to develop and operate large-scale AI models. The investment also demonstrates India’s ambition to become a global hub for AI infrastructure, rather than merely an important market for products developed by U.S. and Chinese companies.

128 Countries Reach Agreement on Autonomous Weapons

128 countries have reached an agreement in Geneva on a text that could pave the way for the first international rules governing autonomous weapons. The negotiations took place under the Convention on Certain Conventional Weapons and followed more than a decade of discussions. However, the text is not legally binding, while the United States and Russia have largely supported the development of national rules rather than the adoption of binding international restrictions. Reuters – States reach agreement at autonomous weapons talks in Geneva

Why it matters: AI could enable military systems to identify, track and select targets with increasingly high levels of autonomy. The agreement represents an important diplomatic step, but it does not yet establish binding rules governing the use of these technologies. The debate has therefore become one of the most important intersections between AI, security and international law: how much autonomy should a machine be given when its decisions can have lethal consequences?

Tesla Cybercab Immediately Comes Under U.S. Investigation

The U.S. National Highway Traffic Safety Administration, NHTSA, has opened an investigation into the Tesla Cybercab, examining how the company determined that the vehicle complies with federal safety standards. The Cybercab is designed for autonomous operation and has an unusual design for a traditional automobile, without the conventional controls associated with a vehicle driven by a human driver. The investigation comes as Tesla begins introducing the vehicle into its autonomous services in the United States. NHTSA – Tesla Cybercab Investigation

Why it matters: The investigation comes at precisely the moment Tesla is attempting to move from demonstrations and testing to commercial operation of driverless vehicles. The case could have implications for the entire autonomous vehicle industry, as regulators must determine how existing safety rules apply to vehicles designed without traditional elements such as steering wheels and pedals. The NHTSA decision could also influence how future generations of robotaxis are approved.

Anthropic Delays IPO, Potentially at a Valuation of Up to $2 Trillion

Anthropic is delaying its plans to go public, with the company expected to begin marketing its IPO no earlier than mid-October. The prospectus is expected toward the end of September, according to sources cited by Reuters. The company’s valuation could reach $2 trillion, while Anthropic is also working on a credit facility of up to $15 billion. Reuters – Anthropic IPO launch shifts toward mid-October

Why it matters: An IPO at such a valuation would represent a major test of investor appetite for frontier AI companies. Anthropic needs significant capital for infrastructure, chips and the development of its models, while a public listing could provide access to additional financial resources. At the same time, a valuation of up to $2 trillion would place enormous pressure on the company to demonstrate that its revenues and growth prospects can justify market expectations.

Sources