In an interview, OpenAI CEO Sam Altman stated the Hugging Face security incident was the first he felt viscerally. He revealed the company paused training to address chained zero-day exploits targeting unreleased models. Altman emphasized the challenge of securing sandboxing and proposed pacing the rate of AI development to allow society time to adapt to new capabilities. The event ranks as his primary concern among top worries.