This was a week about money: Anthropic's IPO filing set the tone, and the stories around it were mostly about who pays for the AI buildout and how.
Anthropic filed its S-1 for a public offering, the week's biggest story. Coverage of the filing describes a fast-growing business, but one carrying a roughly $498 billion valuation asterisk. Commentators also argued that if the stock underperforms after listing, it could prompt a broader reset in how tech values AI companies. Separately, Anthropic launched tools that show companies what their AI spending is actually buying, a practical answer to the cost-justification questions enterprise buyers keep asking.
OpenAI introduced GPT-6.1 Terra, the week's main model release. Meta, meanwhile, launched a plan for a $95 billion stock offering to fund its AI expansion. That is a notable shift in how hyperscale spending is being financed.
Several secondary items round out the picture. Nvidia and AMD will invest $4 billion in Meta, deepening the ties between chip suppliers and their largest customers. Microsoft offered voluntary buyouts to 60,000 employees as AI drives an efficiency push, and reported $10 billion in revenue from OpenAI. Against that, Google and Microsoft both put off raising their capex forecasts, a small sign that the buildout may not accelerate indefinitely.
To watch next week: how the market reads Anthropic's S-1 details, and whether other hyperscalers follow Meta in raising capital through equity.