The last 24 hours in AI were dominated by money and capability. OpenAI quietly handed the SEC the most-anticipated tech S-1 in years, Nvidia delivered another set of generational chip earnings, and Anthropic kept reframing what software engineering looks like with Claude at the center.
OpenAI confidentially files for IPO at up to a $1 trillion valuation
OpenAI confidentially submitted its S-1 to the SEC on Friday, targeting a Q4 2026 public listing at a valuation reported between $852 billion and $1 trillion, with Goldman Sachs and Morgan Stanley leading the deal. The filing comes despite OpenAI losing about $1.22 for every $1 of revenue in Q1 and would, if priced near the top of the range, be one of the largest IPOs ever. Fortune unpacks the questions investors will finally see answered, from unit economics to compute commitments.
Nvidia posts a record $81.6B quarter, guides to $91B
Nvidia reported record fiscal Q1 revenue of $81.6 billion, up 85% year over year, and $58.3 billion in profit, with data-center revenue alone hitting $75.2 billion on the ramp of Blackwell 300 systems. The company guided to roughly $91 billion next quarter and authorized an additional $80 billion buyback. Al Jazeera reports shares slipped slightly after hours as expectations had run even higher.
Anthropic showcases “Code with Claude” and a new dreaming feature
At its two-day Code with Claude event in London, Anthropic argued that Claude is now “as good as a mid-level engineer” and unveiled a feature called dreaming, in which Claude Code agents leave notes for the next agent that picks up the same codebase. Almost half the room said they had shipped a pull request written end-to-end by Claude. MIT Technology Review captures the optimism — and the unease — around what it means for entry-level engineering jobs.
Frontier labs agree to pre-release access for U.S. regulators
Microsoft, Google and xAI have agreed to give the U.S. government early access to frontier models for safety review before public release, a notable shift after the administration spent months pushing back on tougher oversight. The move follows internal alarm over Anthropic’s “Mythos” model demonstrating nation-state-grade cyber-offensive capabilities. The Hill has the details on which agencies will run the evaluations.
What to watch: OpenAI’s roadshow window opens 15 days after the S-1 goes public — the next round of disclosures will shape every other AI deal on the table this summer.