Skip to content
Published:

Major AI Platforms Generated Terrorist Attack Plans During Independent Government Audit

Incoming Chair Hassan Announces AI Safety Legislative Push in New Congress

WASHINGTON — A new official Government Accountability Office investigation requested by U.S. Senator Maggie Hassan (D-NH), incoming Chair of the Senate Homeland Security and Governmental Affairs Committee, found that four of the most-used AI platforms — ChatGPT, Claude, Gemini, and Grok — all provided terrorist attack plans in a real-world test to determine AI safety risks. The results raise serious alarm bells about the need for Congress to take up and pass safety legislation, which incoming Chair Hassan announced today will be an early priority of the Committee.  

“A terrorist can log onto an AI platform and in less than 12 hours walk away with a concrete playbook for how to carry out a deadly attack. Today’s new report makes it clear that the AI public safety threat is no longer theoretical. If Congress does not act, it is a question of when – not if – a bad actor will use AI to plan and carry out a major attack,” said incoming Chair of the Senate Homeland Security and Governmental Affairs Committee, Senator Maggie Hassan. “Congress must quickly take up and pass legislation to address AI safety threats. To that end, in the new Congress, one of the first hearings of the Homeland Security and Governmental Affairs Committee will focus on the risk that AI poses to our homeland security and critical infrastructure, and our committee will prioritize developing legislation to address these threats.”

During an independent government audit, AI platforms were asked to provide plans to conduct either a mass casualty event by developing an improvised explosive device (IED) or to maximize economic harm through an attack on electrical infrastructure in major cities. All four AI platforms provided terrorist attack plans in less than 12 hours, including dates, times, guides to maximize damage, maps, diagrams, and more depending on the platform. While the platforms initially avoided providing the requested information or deleted information after it was provided, all of these safeguards were easily bypassed. Independent experts also confirmed that the details provided by the platforms would, in fact, have detonated explosives or caused harm. For safety reasons, the report withholds the exact prompts used to bypass the safeguards and which platform provided what level of information.

In response to the findings, incoming Chair Hassan announced that she will work to address the threat that AI platforms can pose to public safety and our critical infrastructure, and that one of the Committee’s first hearings in the new year will focus on AI safety. Incoming Chair Hassan is exploring various solutions to address this safety threat, including requiring more rigorous safety testing before AI platforms publicly release models.

Click here to read the full GAO report and see below for an overview of the information that each AI platform provided during the test:

 

AI Platform Test Results

###