Anthropic’s IPO puts AI funding appetite to the test
Anthropic filed paperwork for a public offering that could seek about $100 billion and value the Claude maker at $1.5 trillion to $2 trillion, according to the Wall Street Journal’s filing analysis. It reported $16.3 billion in first-half revenue and an $80.7 billion net loss, largely tied to an accounting charge, the New York Times reported.
The offering could become a major test of investor demand for AI companies as they seek funding for costly computing capacity. The Journal says a strong reception could help rivals including Meta and OpenAI raise capital, while weaker demand could complicate their expansion plans, amid falling AI-related stocks and tighter financing conditions according to its market analysis. Anthropic’s filing also shows adoption remains early at many businesses the Journal noted. (NY Times, NY Times, WSJ, Transformer AI, WSJ, NY Times)
OpenAI expands GPT-6.1 options and cloud access
OpenAI introduced GPT-6.1 Terra, a mid-priced model designed for visual reasoning, computer use and clearer writing. Simon Willison reports it costs $1 per million input tokens and $5 per million output tokens—about half the price of GPT-6.1 Sol, though it is less capable on many other tasks. OpenAI describes it as faster than its predecessor, GPT-5.6 Terra.
Separately, Ars Technica reports that Google Cloud plans to offer OpenAI’s GPT-6 and 6.1 models, including Astra, Sol and Luna, through Model Garden and its marketplace. Businesses will be able to run them on Google infrastructure and shift workloads among models. The moves add a more affordable model option and broaden where companies can access OpenAI systems. (Simon Willison, Ars Technica)
Broadcom helps secure record $63 billion Anthropic financing
Broadcom CEO Hock Tan stepped in to secure a $63 billion financing package for AI chips to be leased by Anthropic, agreeing to higher loan rates and stronger lender protections, the Wall Street Journal reported. The deal was announced after Anthropic filed for an initial public offering, according to the financing report.
The package brings together bank loans and private credit to buy Broadcom- and Google-designed computing equipment, with deliveries expected to begin next year. The Journal’s account of the negotiations says banks had grown wary of AI-sector risk and rising rates; Tan also agreed to guarantees against some losses, the deal report said. The financing underscores the scale of capital sought by AI companies as Anthropic prepares for a possible IPO; the IPO report cited a potential valuation of $1.5 trillion to $2 trillion.
Quote of the Day
I think the cascading hierarchy or critical path is now very much the network. And I think people have grossly underestimated the amount of infrastructure that’s going to be required.
Ciena CEO Gary Smith, WSJ
Other Links
- NY Times: In the A.I. ‘Robogym,’ Machines Train to Be More Like Humans
- NY Times: U.S. Pensions Are Riding the A.I. Wave. For Better or Worse.
- WSJ: Nvidia to Invest $3 Billion in Meta as Facebook Parent Seeks Up to $95 Billion
- The Stanford Daily: Anthropic illegally trained Claude with Stanford professors’ books
- Google DeepMind: Genie 4: A new generation of world simulation — Google DeepMind
- Gizmodo: Police Are Using AI to Pull Over Drivers for ‘Suspicious’ Behavior
- WSJ: AI Build-Out Fuels Surge in Demand for Outdoor Storage
- WSJ: ASM International Raises Sales Forecasts Amid Soaring AI Demand
- NY Times: Microsoft Offers Buyouts to 60,000 Employees as It Spends on A.I.
- WSJ: China Industrial Profits Rose 14.4% in September Led by High-Tech Surge