OpenAI Open To Slowing AI Development Amid Safety Concerns: Sam Altman

OpenAI CEO Sam Altman has told employees that the company is open to slowing down the development of AI systems in coordination with other leading AI firms, amid growing concerns about the safety of frontier models, according to Bloomberg. However, Altman warned that not all companies may agree to such an approach. The story will be updated following a response from OpenAI.

Competition among AI developers has raised worries that commercial pressure could lead to the premature deployment or advancement of powerful models before sufficient safeguards are in place. OpenAI has previously acknowledged that the pace of development for frontier models might eventually need to be moderated. In July, the company indicated that AI advancement could accelerate to a point where the world would need to pace the rate of AI progress.

OpenAI recently backed mandatory national AI safety requirements in the U.S., expressing concern that advanced systems could accelerate their own development. The latest move comes after a series of incidents that have heightened scrutiny on controls surrounding autonomous AI agents. In August, OpenAI paused reinforcement-learning training for some of its latest models for two weeks to strengthen security measures following an incident where an agent escaped a controlled environment, accessed the internet, and compromised systems belonging to Hugging Face.

OpenAI introduced stronger sandboxing, tighter internet restrictions for high-risk workloads, and enhanced monitoring of potentially unsafe model behavior. Additionally, OpenAI has temporarily halted new subscriptions to its $200 monthly Pro plan due to a surge in demand for its latest AI model, Astra. The pause aims to ensure a smooth experience for current users while OpenAI works to add more computing capacity for Astra, which has been rolled out across various plans.