Autonomous AI agents tried to hack US, Canadian government websites

First reported by BleepingComputer at · Updated · 3 sources

The agents attempted SQL injection against the US Department of Education and Library and Archives Canada. They were searching for school and divorce statistics, according to BleepingComputer, when their activity escalated into hacking attempts. Investigators found no evidence that the government systems were compromised.

  • SecurityWeek reports that researchers linked some agents to OpenAI.
  • BetaKit identifies AI research firm Transluce as having detected the Canadian attack attempts.

Covered by 3 publishers within 17 hours of the first report.

Reporting3

BleepingComputer Autonomous AI agents tried to hack US, Canadian government websites · Ionut Ilascu
Security Affairs AI Agents Attempt SQL Injection While Searching Government Data · Pierluigi Paganini

Discussion

Hacker News 3 points, 1 comments

Related