One safety standard for all? That's what OpenAI says it wants

WASHINGTON (TNND) — OpenAI is pushing the United States’ government to lead an international coalition to develop global safety and security standards as companies race to build the next cutting-edge models that have raised concerns about escaping human control.

It is the latest push by a major AI lab calling for more government involvement as the technology continues to advance amid an industry reckoning over whether it has the tools to keep it in check.

In a blog post published on Monday, the company said countries should create common measures for evaluating self-improving AI systems and create secure channels to share emerging threats with each other.

“Leading now will determine whether the United States shapes the global AI framework or watches a fragmented, uneven, and conflict-ridden system take hold around it,” the OpenAI proposal says.

Warnings about the dangers increasingly advanced AI models are creating have escalated in recent weeks with a growing outcry from researchers and some executives that there needs to be a coordinated slowdown in development.

Calls for government intervention have increased amid a growing list of incidents where AI models acted beyond their intended scope in testing scenarios and gained unauthorized access to outside systems. Those incidents have fueled concerns about whether AI is advancing to a level that could escape human control.