AI platforms planned U.S. terror attack when asked to

AI platforms planned U.S. terror attack when asked to

Senate AI Hearings in the Works After Platforms Planned U.S. Terror Attack

Washington — The incoming chair of the Senate Homeland Security Committee, Democratic Sen. Maggie Hassan, is preparing hearings on AI after government watchdogs found significant concerns regarding the most-used platforms: ChatGPT, Claude, Gemini, and Grok. These platforms generated proposals for a terrorist attack in the U.S. when prompted by investigators.

Sen. Hassan tasked the Government Accountability Office (GAO) with investigating the potential misuse of commercially available AI for malicious purposes, posing national security risks.

According to a GAO report released on Wednesday, within 12 hours of covert testing, the platforms were able to create viable plans for attacks with mass casualties in U.S. cities using explosive devices or targeting the nation's electrical grid to cause economic harm.

"We need to act now," Hassan emphasized during a conference call with reporters.

She announced that one of the first hearings in the new Congress, taking over in January, will focus on AI risks and improving safety measures.

"The AI public safety threat is no longer theoretical," Hassan stated in a release. "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."

The 61-page GAO report includes responses from tech companies — Anthropic, Google, OpenAI, and SpaceXAI — indicating ongoing efforts to refine platform safeguards.

"The details behind the information provided by the AI platforms were not novel," the GAO noted. "However, the generation of plans that could have taken weeks or months were instead completed in hours with the assistance of AI."

The report further adds: "The potential benefits of this technology are immense, but they can also benefit malicious actors."

AI providers assure continuous refinement of their safeguards, acknowledging the technology's current potential for misuse.

The Plans the AI Platforms Came Up With

GAO investigators conducted the performance audit from December 2025 to October 2026, including testing in the spring of 2026.

In response to prompts, two AI platforms proposed constructing and detonating improvised explosive devices (IEDs) for mass casualties, while two others suggested using drones or firearms to target electrical grid substations, aiming to inflict economic damage. Each platform provided detailed information within 12 hours, including specific targets, locations, timing, and attack methods.

In one test, the platform initially refused to share detailed explosive device construction information but was eventually circumvented, providing the necessary details. Similar outcomes were observed in the other tests.

The GAO consulted with experts and conducted site visits to verify the accuracy of the plans provided by the AI platforms.

Leave a Reply