Tech companies signed a voluntary AI safety agreement with the White House
Google, OpenAI, Meta, and others promised internal monitoring, though the pact lacks the weight of formal federal regulation.
Top executives from Google, Anthropic, Meta, OpenAI, xAI, and Nvidia gathered at the White House for a luncheon with President Donald Trump. They signed an agreement titled The White House Accord on Super Intelligence.
The document outlines four main suggestions for AI labs. Companies are encouraged to build internal systems to monitor model performance and prevent unintended behaviors. They should assign internal teams to handle safety remediation. The accord also recommends using external monitors for independent assessments and creating board-level committees to oversee these efforts.
This is the kind of pledge that the FTC could potentially enforce, if a company materially failed to follow through on any of these promises.
While this pact signals a focus on safety, it is not a binding law. Failing to follow these steps could be viewed as misleading consumers, potentially triggering an investigation by the FTC. However, legal experts note that enforcement would likely focus on deceptive practices rather than broad structural reforms.
This is not the first time these companies have participated in voluntary agreements. Earlier in 2025, similar commitments regarding red-teaming and information sharing were established in the United Kingdom and the Republic of Korea.
Reports suggest the FTC is considering a broad inquiry into several frontier labs and the evaluation group METR. Agency leaders have previously stated they are searching for evidence of risks, though it remains to be seen if this will lead to concrete regulatory action.
Douglas Farrar, a former FTC official, expressed skepticism about the impact of the new accord given the current administration's stance toward the industry. Do you believe voluntary industry pledges are enough to manage the risks of advanced AI?
Source: WIRED: Artificial Intelligence