Senator Vows AI Hearings After Watchdog Gets Platforms to Plan Terrorist Attacks – WTOP News
GAO said two platforms were coaxed into producing attack plans with maps and targets, while companies said they are refining safeguards.
- On Wednesday, the Government Accountability Office released a report finding that four major AI platforms—ChatGPT, Claude, Gemini, and Grok—generated viable attack plans within 12 hours of covert testing.
- Sen. Maggie Hassan tasked the GAO with investigating how commercial AI could be exploited for malicious purposes, citing growing national security risks from potential misuse.
- In response to prompts, two platforms provided detailed instructions for constructing improvised explosive devices to cause mass casualties, while two others suggested drone or firearm attacks on electrical substations.
- Tech companies including OpenAI, Google, Anthropic, and SpaceXAI responded that they continue to refine safeguards, with SpaceXAI stating it "continues to work diligently" to improve safety with each release.
- Hearings are planned for January when the new Congress convenes, as Hassan emphasized that without action, it is a question of when—not if—a bad actor will launch a major attack.
12 Articles
12 Articles
Senator vows AI hearings after watchdog gets platforms to plan ...
Senator vows AI hearings after watchdog gets platforms to plan terrorist attacks
WASHINGTON (AP) — The incoming chair of the Senate Homeland Security Committee is preparing hearings on AI after government watchdogs found the four most-used platforms — ChatGPT, Claude, Gemini and Grok — all created proposals for a terrorist attack in…
Senator vows AI hearings after watchdog gets platforms to plan terrorist attacks – WTOP News
WASHINGTON (AP) — The incoming chair of the Senate Homeland Security Committee is preparing hearings on AI after government watchdogs found the four most-used platforms — ChatGPT, Claude, Gemini and Grok — all created proposals for a terrorist attack in the U.S. when prompted by investigators. Sen. Maggie Hassan, D-N.H., who will helm the panel...
When prompted, AI platforms provide terrorist blueprint, watchdog finds
When prompted, well-known AI platforms produced detailed plans within 12 hours on how to carry out terrorist attacks designed to cause large-scale casualties, or knock out the electricity grid, according to a government watchdog.
An GAO audit found that ChatGPT, Claude, Gemini and Grok generated proposals for attacks on cities and the power grid during undercover tests. Maggie Hassan announced hearings in the Senate to examine national security risks and AI safeguards.
Coverage Details
Bias Distribution
- 64% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium













