Skip to content

Senator Plans AI Hearings After Platforms Generate Terrorist Attack Plans

The GAO found that two platforms provided detailed instructions for constructing improvised explosive devices, while the other two suggested attacks involving drones or firearms targeting electrical substations.

Senator vows AI hearings after watchdog gets platforms to plan terrorist attacks. Pic via(@CTVNews)

A Government Accountability Office report has raised concerns about the potential misuse of artificial intelligence after four major platforms generated viable attack plans during covert testing. The report, released Wednesday, examined ChatGPT, Claude, Gemini and Grok.

The GAO found that two platforms provided detailed instructions for constructing improvised explosive devices, while the other two suggested attacks involving drones or firearms targeting electrical substations.

💡
Sen. Maggie Hassan, who requested the investigation, said the findings highlight growing national security risks associated with commercial AI systems. OpenAI, Google, Anthropic and SpaceXAI said they continue working to strengthen their safety safeguards.

Hassan plans to push for congressional hearings in January, when the new Congress convenes. She warned that without stronger action, a major attack by a malicious actor could become increasingly difficult to prevent.

Related Tweet:

Also Read:

‘INEXCUSABLE’: Lawmaker Slams Commerce Dept For Inaction
Despite these concerns, the Commerce Department has not placed Megaspeed on its Entity List of companies subject to trade restrictions.

Comments

Latest