OpenAI president says company has already slowed cutting-edge AI development

Brockman told Bloomberg's "Odd Lots" podcast that the company is pulling back earlier in its development cycle to focus on alignment training. He also shared in an interview with Andreessen Horowitz that OpenAI paused some projects and directed its Astra model to probe its own infrastructure for critical vulnerabilities. "We took 25% of our production engineers and said, 'Sorry, all your projects are on hold. You are now defending. You are now up-leveling our security architecture,'" Brockman said.

This initiative led to the discovery and fixing of several serious issues, including priority zero problems, which are the most urgent. Brockman noted that this process would need to be repeated for every new model. These comments come amid a broader industry debate about the pace of AI advancement, following a researcher's warning that leading AI companies are "gambling with our lives."