A nascent effort to develop technology that could verify a global slowdown or pause in artificial intelligence development is underway, with companies like Anthropic dedicating resources to create these verification systems. Anthropic stated in a June blog post that such systems would "enable frontier AI developers to verify that others globally have actually stopped or slowed, and that a bad actor could not use the auspices of a coordinated slowdown to jump ahead in secret." The company has announced it would devote resources to building the systems that a credible slowdown or pause would require.
Discussions have occurred between some technology firms and governments regarding this work, though specifics about the governments involved and the nature of the discussions remain undisclosed. For instance, Amodo has held preliminary discussions with governments about its work, according to an individual familiar with the matter. However, the U.S. government has expressed a different perspective.
Michael Kratsios, director of the White House Office of Science and Technology Policy, stated in February that the U.S. "totally reject[s] global governance of AI," adding, "We believe AI adoption cannot lead to a brighter future if it is subject to bureaucracies and centralized control."
While the White House has slightly adjusted its regulatory approach since February, reportedly influenced by the cyber warfare capabilities of recent AI models and now testing some frontier models before release, skepticism toward significant interventions persists. White House officials remain concerned that heavy-handed regulations could cede ground to China. President Trump, in the introduction to a June executive order, wrote, "We refuse to stifle [AI] innovation with overly burdensome regulation."