Major AI industry figures who supported Donald Trump’s campaign for deregulation now advocate for a formal AI oversight framework following what they characterize as chaotic and unpredictable government interventions. The shift in stance stems from recent administrative actions, including export controls and model launch restrictions, which executives claim have created a de facto licensing regime. While industry leaders previously campaigned against the Biden administration’s AI policies, many now report that current, ad-hoc oversight is more disruptive to their operations.

The administration’s approach began with a June 2 executive order implementing a voluntary 30-day model review process. However, this framework was quickly bypassed when the White House imposed export controls on Anthropic’s Mythos 5 and Fable 5 models on June 12 after security concerns were raised. Furthermore, the administration pressured OpenAI to limit the launch of its model, Sol, to 20 government-approved partners. This marked the first instance of a frontier model being released under a government-managed access list.

Policy experts, such as Saif Khan, a former Commerce Department adviser, describe the current strategy as opaque and lacking the necessary infrastructure or expertise to manage risks effectively. The resulting uncertainty has effectively created a moratorium on new product releases that may impact company finances. Dean Ball, a former Trump official joining OpenAI on July 6, noted that while government security concerns are legitimate, the administration’s execution has been flawed. Industry groups, including the Software and Information Industry Association, are now coordinating a push to replace these reactive measures with a codified, predictable regulatory framework. White House spokesperson Liz Huston stated that the administration remains focused on ensuring American dominance in AI while continuing efforts to preempt state-level regulations.

Major AI industry figures who supported Donald Trump’s campaign for deregulation now advocate for a formal AI oversight framework following what they characterize as chaotic and unpredictable government interventions. The shift in stance stems from recent administrative actions, including export controls and model launch restrictions, which executives claim have created a de facto licensing regime. While industry leaders previously campaigned against the Biden administration’s AI policies, many now report that current, ad-hoc oversight is more disruptive to their operations.

The administration’s approach began with a June 2 executive order implementing a voluntary 30-day model review process. However, this framework was quickly bypassed when the White House imposed export controls on Anthropic’s Mythos 5 and Fable 5 models on June 12 after security concerns were raised. Furthermore, the administration pressured OpenAI to limit the launch of its model, Sol, to 20 government-approved partners. This marked the first instance of a frontier model being released under a government-managed access list.

Policy experts, such as Saif Khan, a former Commerce Department adviser, describe the current strategy as opaque and lacking the necessary infrastructure or expertise to manage risks effectively. The resulting uncertainty has effectively created a moratorium on new product releases that may impact company finances. Dean Ball, a former Trump official joining OpenAI on July 6, noted that while government security concerns are legitimate, the administration’s execution has been flawed. Industry groups, including the Software and Information Industry Association, are now coordinating a push to replace these reactive measures with a codified, predictable regulatory framework. White House spokesperson Liz Huston stated that the administration remains focused on ensuring American dominance in AI while continuing efforts to preempt state-level regulations.