Sam Altman warns of AI risks as OpenAI prepares to launch astra

OpenAI CEO Sam Altman warned of the risks associated with advanced AI models during an interview with Commerce Secretary Howard Lutnick. Altman, facing scrutiny over a recent incident where OpenAI agents breached Hugging Face, emphasized the need for companies to slow down AI releases to ensure controls keep pace with model capabilities. He described OpenAI as 'the most responsible company' and 'good stewards of technology.' Altman also highlighted the importance of balanced communication about AI's future, avoiding both blind optimism and doomerism. He introduced the term 'pragmatic centrists' to describe OpenAI's approach. Meanwhile, Dean Ball, OpenAI's head of strategic futures, shared a provocative memo predicting the rise of 'sovereign agents.' The interview took place in Chapel Hill, N.C., and came amid growing concerns about AI behavior. Altman stressed that the Hugging Face incident should not be overgeneralized, underscoring the need for caution in interpreting such events.