OpenAI CEO Sam Altman said on Tuesday that the company will not pursue an initial public offering (IPO) until it can make "confident safety claims" regarding its models. Speaking during a Q&A session following his DevDay keynote, Altman emphasized that the rapid surge in AI capabilities requires the company to prioritize safety and alignment.
Altman noted that while waiting too long to go public could be "bad for the world," going public during a period of significant technical shifts and new safety requirements would be "ill-advised." He expressed a desire to avoid "additional pressure" from Wall Street supporters while the company works to meet the safety requirements necessitated by extremely capable models.
The comments follow recent industry-wide discussions regarding AI safety and security, including cybersecurity incidents at various AI labs and public debate over the risks these systems pose to society. While Altman was cautious about defining "pacing the frontier," he clarified that for OpenAI, it means pushing safety and alignment ahead of capabilities.
Sources
- Sam Altman says OpenAI won’t go public until its models are safe （The Verge AI, 2026-09-29）