Congressional watchdog gets AI chatbots to plan terror attacks against America in alarming test

Congressional watchdog gets AI chatbots to plan terror attacks against America in alarming test

AI chatbots helped plan terrorist attacks, including guides on how to maximize damage, as part of a government test that has raised “serious alarm bells” about the technology’s rapid advancement and the need for legislation.The test was carried out in May 2026 by the Government Accountability Office at the request of Democratic Sen. Maggie Hassan of New Hampshire, the incoming chair of the Senate Homeland Security and Governmental Affairs Committee. Hassan said the watchdog’s findings, released in a report Wednesday, made clear “that the AI public safety threat is no longer theoretical.” It comes amid heightened concern surrounding AI following major outages affecting air traffic control, online banking, and other personal data in the U.S. this week.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, a news release from Hassan said. In the test, the platforms were asked to provide plans to conduct either a mass-casualty event by developing an improvised explosive device or to maximize economic harm through an attack on electrical infrastructure in major cities, the report said.AI chatbots helped plan terrorist attacks, including guides on how to maximize damage, as part of a government test that has raised ‘serious alarm bells’ about the technology’s rapid advancement (AFP/Getty)One platform, which was not identified in the report, recommended a specific date and time of attack “because several public events were scheduled in the city and large crowds were expected to attend,” it said. Two platforms suggested placing a bomb in a vehicle and one in a hotel parking garage as an attack method. The report found that the platforms provided plans in “less than 12 hours, including dates, times, guides to maximize damage, maps, diagrams, and more depending on the platform.” The chatbots initially refused the request, but eventually generated the plans for a terrorist attack, the report said.“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,” Hassan said. “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.”Hassan said that the new Congress “will focus on the risk that AI poses to our homeland security” when it convenes in January 2027.The AI companies behind the platforms - Google, Anthropic, OpenAI and SpaceXAI- said that their systems have been strengthened since the spring, when the government’s test was carried out. The tech giants said that they “take seriously the risk that AI could be used to facilitate violence, as well as the responsibility to protect people and communities from harm,” the government’s report summarized.Two platforms suggested placing a bomb in a vehicle and one in a hotel parking garage as an attack method, the report found (pictured) (Government Accountability Office)The report was released as Homeland Security Secretary Markwayne Mullin is set to appear before the House Committee on Homeland Security Wednesday to answer questions about Friday’s rare nationwide ground stop. His department oversees the Cybersecurity and Infrastructure Security Agency. Mullin and Transportation Secretary Sean Duffy have ruled out a cyberattack and said there was nothing nefarious about the incident. The Trump administration has blamed “old equipment” after the outage appeared to be caused by work to replace old copper wire with fiber cables at a site in Arkansas. The Department of Homeland Security chief raised eyebrows Monday night on Fox News, where he suggested to Sean Hannity that the disruptions were related to AI - which he referred to as the Trump-preferred “SI” throughout the interview. Mullin’s source for the theory was “one guy who is very smart with tech,” whom the secretary kept anonymous. After watching the segment, Trump administration group chats lit up with “WTFs,” according to Politico.

Original Source

Read the full article at Independent →

KhanList aggregates and links to publicly available news content. We do not host full articles from third-party sources. Always verify important information with original sources.