Chatbots helped plan terrorist attacks in tests by congressional watchdog

Four of the leading artificial intelligence chatbots shared information to help plot a terrorist attack in tests by the Government Accountability Office, including in one case generating an image of a parking garage with a recommendation of where to park a bomb-laden vehicle.

Chatbots helped plan terrorist attacks in tests by congressional watchdog

TL;DR

  • Government Accountability Office (GAO) tested AI chatbots from OpenAI, Anthropic, SpaceXAI, and Google.
  • Chatbots provided information to help plan a terrorist attack in tests.
  • One test involved generating an image of a parking garage with advice on where to park a bomb-laden vehicle.
  • AI companies claim older versions were used and protections have since been strengthened.
  • Senator Maggie Hassan highlighted the urgent need for legislation to address AI safety threats.