WASHINGTON, DC -- Stakeholders from industry, government, academia, and civil society are invited to take part in new standards work focused on managing the unique risks of frontier AI.

The work centers on the most advanced general-purpose AI systems—technologies whose accelerating capabilities may introduce distinct risks to public safety and security, including the potential creation of advanced cybersecurity or biological threats.

A new consensus-based effort will develop a specialized, fit-for-purpose framework to help address those concerns across the full risk management lifecycle. The work is being carried forward through the InterNational Committee for Information Technology Standards (INCITS), an ANSI-accredited standards developer, as project INCITS 594‑202x, Information Technology—Framework for Managing Unique Risks from Frontier AI.

“Standardization is the bridge between breakthrough innovation and public trust,” said Rohit Israni, chair of INCITS/Artificial Intelligence and CEO of CertientAI. “As we move into an era of increasingly autonomous and capable systems, this framework will provide the essential guardrails needed to navigate emergent risks unique to frontier AI.”

“Frontier AI holds enormous promise, but it also poses risks that need to be managed responsibly,” said Chris Meserole, executive director of the Frontier Model Forum. “This standard will ensure that frontier AI risk management keeps pace with the technology itself.”

Technologists, researchers, policymakers, and other stakeholders interested in contributing can contact INCITS at incits@itic.org.

Hear More at the 2026 ANSI Innovation Summit  
The conversation continues at the 2026 ANSI Innovation Summit, July 28–31 at the Grand Hyatt Denver. On Wednesday, July 29, the Summit’s conference on AI, quantum, and the digital economy features remarks from Rohit Israni, followed by a morning session devoted to this very topic. Moderated by Chris Meserole, Managing Unique Risks from Frontier AI will convene experts from Microsoft, NIST, and UC Berkeley’s Center for Long-Term Cybersecurity for a discussion on how a fit-for-purpose framework can support the responsible development of the most advanced AI systems.

Registration is now open. Visit www.ansi.org/summit to learn more and reserve your spot.