National/International News

Investigators claim AI has planned terrorist attacks

Getting your Trinity Audio player ready...

by the Journal-Opinion

“Four of the most widely used artificial intelligence platforms generated viable plans for attacks on U.S. critical infrastructure and mass-casualty events during covert government testing, despite safeguards designed to prevent their misuse,” reported Defense One. 

Investigators tested the ChatGPT, Grok, Gemini, and Claude platforms in the spring. 

“They found that all four platforms provided information that could help malicious actors plan and carry out attacks.”

The findings were disclosed in a Government Accountability Office report published on Wednesday. The federal office conducted the investigation at the request of U.S. Sen. Maggie Hassan, D-New Hampshire. 

In a news release, she called on Congress to pass new legislation to address AI safety threats.

“GAO did not issue recommendations in the report, which examined third-party products rather than its typical examination of federal agencies. Instead,” Defense One reported. “GAO said it sought to highlight the risks for lawmakers and bring attention to safety concerns.” 

The chart below from the GAO report was sent out by Hassan’s office.  


Discover more from Vermont Daily Chronicle

Subscribe to get the latest posts sent to your email.

All topics and opinions welcome! No mocking or personal criticism of other commenters. No profanity, explicitly racist or sexist language allowed. Real, full names are now required. All comments without real full names will be unapproved or trashed.