OpenAI CEO Sam Altman calls safety a condition for IPO — The Verge

OpenAI CEO Sam Altman said the company will not go public until it can confidently state that its artificial intelligence models are safe. He said this while answering journalists’ questions after the keynote speech at the DevDay conference, The Verge reports.

According to Altman, OpenAI intends to continue developing AI, but the rapid growth in model capabilities requires convincing statements about their safety. He did not name specific timeframes for a possible IPO. At the same time, Altman believes that delaying an IPO for too long could also be negative for the world.

Safety and investor pressure

Altman noted that public status could create additional pressure on OpenAI from Wall Street investors. In his view, going public during the transition to significantly more powerful models and the emergence of new safety requirements would be inappropriate.

More current news is available on the UA.News Telegram channel Telegram.

He explained that by a cautious pace of developing advanced AI systems, the company means a situation in which safety measures and model alignment outpace the growth of their capabilities. According to him, OpenAI seeks to scale the next stage of AI development only when it can confidently speak about the safety of such systems.

Debate over control of AI

Altman’s comments came amid disputes over whether OpenAI and other developers are capable of controlling their own systems. According to information that emerged in July, an unreleased OpenAI model hacked competitor Hugging Face’s lab without the company’s knowledge. Other cybersecurity incidents at OpenAI, Anthropic, Meta, and Google later also became known.

Earlier this month, Altman also said in an interview that OpenAI would likely not go public this year. OpenAI competitor Anthropic officially filed documents to go public in June. According to the outlet, Anthropic’s IPO could take place in November after the U.S. midterm elections.