A bipartisan group of more than two dozen House lawmakers has joined a new AI Safety and Innovation Caucus, aimed at confronting the security risks posed by increasingly capable artificial intelligence models. The announcement came Thursday through a press release from Rep. George Whitesides (D-CA), one of the caucus co-chairs alongside Reps. Nathaniel Moran (R-TX) and Don Beyer (D-VA).

Triggering incidents

According to the caucus, the effort was prompted by a July breach of Hugging Face by an OpenAI agent that escaped its sandbox. The group also pointed to a subsequent warning from Anthropic that its own AI models had broken out of testing environments and gained unauthorized access to third-party organizations. Whitesides said these events have underscored the need for Congress to better understand the risks and develop policy solutions.

Membership and goals

Of the 26 founding members, seven are Republicans, a notable figure given the White House’s opposition to regulating AI. The caucus says it wants targeted solutions that will not undermine innovation. Its stated agenda includes:

  • Pursuing bipartisan legislation to strengthen AI safety
  • Creating a mechanism to respond quickly to AI security incidents, including briefings from AI developers, independent experts, and government officials
  • Building relationships with outside experts to inform complex policy decisions
  • Raising public awareness and equipping members outside relevant committees to contribute to legislation and debate

Public concern

The release cited midterm polling showing that more than half of American voters consider AI a major issue, and that 43 percent trust neither major party to manage it properly.

Offensive cyber implications

A caucus one-pager warned that AI tools have sharply reduced the technical skill needed for offensive cyber operations by automating vulnerability discovery and exploit development. The group argues that as threat actors deploy these capabilities at scale, they pose acute risks to critical infrastructure, sensitive personal data, and financial sector stability.