OpenAI to Expand AI Access for Critical Infrastructure

OpenAI President Greg Brockman is expected to announce plans this Thursday at a cybersecurity summit at the company's headquarters to expand access to OpenAI tools for critical infrastructure and public sector agencies. About 300 security chiefs and CISOs from Fortune 100 companies are expected to attend.

This follows an accidental intrusion into Hugging Face and recent cyberattacks on U.S. water systems suspected to be AI-powered, raising concerns about the resilience of utilities against future AI-enabled threats. The summit will focus on how AI tools can better support defenders, including those protecting critical infrastructure and public services.

The announcement comes as OpenAI addresses safety concerns about its models. After the Hugging Face breach, OpenAI's upcoming Astra model has reached a "critical" level of cybersecurity capability. Additionally, OpenAI is researching new training methods, raising worries among security professionals about transparency into model reasoning. Last week, OpenAI publicly warned that organizations have only months to prepare for AI-powered cyberattacks.