OpenAI is considering delaying its IPO to next year, per a New York Times report covered in this week's market wrap. The cited reasoning is SpaceX's weak post-IPO performance (SPCX has retreated from its day-one $160 close) and the broader AI-share volatility that drove the Nasdaq's fifth losing session Friday. Both OpenAI and Anthropic filed confidential S-1 forms in June at near-trillion-dollar private valuations. OpenAI just blinked first.
The dynamic the delay confirms is the one we flagged when both companies filed: two $1 trillion-class AI listings competing for institutional allocations at the same time creates pricing pressure that helps neither company, and the loser of that timing race is whichever one prices second into a softer market. SpaceX's IPO was the largest in history at roughly $75 billion raised at $135, and the day-one pop to $160.95 looked like a clean win until the stock retreated below its open inside two weeks. That tape is what OpenAI's bankers are now staring at.
For the cap-table reading, the delay is rational for OpenAI specifically. A 2027 listing window gives them another year of revenue ramp to compress the trailing multiple at the $1 trillion mark, and it lets the public market digest the AI-equity supply that already hit (SpaceX) and the supply still pending (Anthropic, SoftBank's Roze, whoever else). Anthropic now has to decide whether to follow the delay or price ahead of OpenAI to claim first-mover positioning. The downside of going second has just gotten more visible.
Bottom Line
The post-SpaceX tape made $1 trillion-class AI IPOs look harder than they did three weeks ago, and OpenAI is responding by buying time. Whether Anthropic follows the delay or races to price first is now the calendar question for AI exits in 2026.