While drawing a line at catastrophic outcomes like a total loss of human control, Altman said there is "a lot of daylight" between his views and those of Anthropic regarding restrictive regulation. He dismissed President Trump’s new AI accord as having "no real power" and predicted a "tidal wave" of cybersecurity issues arising from open source models. Despite these hazards, Altman argues the benefits ofbroad individual agency and access outweigh the risks of failures that do not threaten humanity's survival.
Tuesday, October 20, 2026
Story
OpenAI Delays Astra 6.1 for Safety and Rejects Lab Gatekeeping
AI16 days ago
Sam Altman believes the world must tolerate certain errors and misuse to ensure that artificial intelligence remains broadly accessible. In an interview for Politico's Decoded newsletter, the OpenAI CEO rejected the notion of "AI gatekeeping," calling the idea that a single lab should control the technology's distribution a "completely unacceptable trade-off." Altman noted that OpenAI had already applied this internal risk judgment by delaying the release of Astra 6.1, as the model was ready for deployment but posed too high an accident risk.
Looking for more context or analysis?
Tilt chat provides reliable, tailored insights, alerts, and automations.