Abstract:
The Information reported, citing people familiar with the matter, that Google, OpenAI and Anthropic are advancing a plan to establish an artificial intelligence (AI) safety standards agency that will not be subject to government supervision. The goal is to launch before the end of this year or early 2027.
The tentative plan is to name this self-regulatory organization the "Standards Authority for Frontier AI".
Working group members have approached several "high-profile" individuals to serve as CEOs.
The contacts include Sriram Krishnan, a former venture capitalist and former AI policy adviser to the Trump administration.
Arati Prabhakar, a former Biden administration official, was also considered.
The list of candidates for chairman includes former US Secretary of State Condoleezza Rice and entrepreneur David Friedberg.
The report added that some people on the shortlist said that no one had yet contacted them.
The standards body will support third-party organizations in testing models before deployment and develop specifications for how developers should report security and security incidents.
The organization is still discussing whether it should conduct its own safety and capability testing of AI models.
According to reports, the White House has previously told Google, Anthropic and OpenAI that companies must first reach an industry consensus on standards work before the government will participate.
Comments