Sam Altman says it may be time to pace AI development

OpenAI CEO Sam Altman recently said it may be time to “pace the rate of AI development” so that society can adjust to new capability levels. On the latest episode of TechCrunch’s Equity podcast, hosts Kirsten Korosec, Sean O’Kane, and Anthony Ha discussed how Altman’s comments were probably prompted by a recent hack in which an OpenAI agent breached Hugging Face’s systems.

O’Kane noted that while a hack carried out by an AI agent is novel, the hack itself wasn’t especially advanced. He compared it to Nixon aides breaking into the Watergate complex rather than a stealthy cyber operation, since the agent was not instructed to hide its tracks and did not need to. Researchers who examined the breach reportedly concluded it was not especially sophisticated, and that the model behaved in a loud, human-like way rather than covering its steps. Ha added that the breach appeared to stem from human error as much as AI capability: the agent should not have been able to get online in the first place, but the company hosting the test apparently failed to secure it properly.

Korosec pointed out that while Altman chose his words carefully — he called for “pacing” development rather than a pause — OpenAI and Anthropic both backed a petition reflecting the same call for caution, and that the Hugging Face incident probably alarmed both Altman and much of the industry. She questioned whether OpenAI can reconcile that caution with its need to keep generating revenue, raising money, and preparing for a possible IPO.

Ha, meanwhile, questioned whether “acceleration versus deceleration” is even the right framework, arguing it implies there is only one path and that the only choice is how fast to move along it, rather than considering different guardrails or approaches entirely. He also noted that Altman has floated 2027 for an initial public offering and that OpenAI has only filed a confidential filing to keep that option open — giving the company more room to talk about pacing than Anthropic, which is already in conversations with bankers and headed toward a nearer-term IPO.

Why it matters

The debate signals that AI labs’ public rhetoric about safety and pacing is increasingly shaped by market timing, with OpenAI able to talk caution more freely than Anthropic, which is closer to an IPO.