AI agent attempted to hack Library and Archives Canada, says U.S. research group
The website for Library and Archives Canada was one of the targets of a failed hacking attempt by an artificial intelligence agent, a San Francisco-based lab that monitors AI activity said in a report this week.
Transluce posted an incident report Wednesday detailing what it describes as “aggressive attempts” to access public data on Canadian and U.S. government web pages.
The report says the attempted hacks occurred on May 28 and June 9 and involved 899 search queries into divorce records between 1905 and 1911.
Thirteen of the requests involved “attack payloads” seeking vulnerabilities in the search parameters, the report says.
Transluce said it does not “confidently attribute” the attempt to OpenAI-powered agents but notes this episode was consistent with similar incidents attributed to OpenAI in a similar time frame.
The report said the incident was reported to the Canadian government on Monday. The Canadian Centre for Cyber Security issued a statement Tuesday saying it’s aware of possible suspicious AI agent activity but there is no indication systems were breached.
That government statement does not mention the nature of the attempted hack but notes “public-facing government websites routinely receive automated and potentially malicious requests.”
Artificial Intelligence Minister Evan Solomon said in online post Tuesday evening the government is working with the Canadian Centre for Cyber Security.
“Protecting Canadians and safeguarding government systems remains a top priority,” he said.
Reports of rogue AI agents have helped fuel widespread public alarm about AI safety in recent weeks.
OpenAI disclosed over the summer that its AI agents hacked AI startup Hugging Face without being instructed by any human to do so.
Other disclosures followed. Anthropic said its artificial intelligence models hacked into three other organizations during testing. In August, Meta disclosed that one of its AI models accessed the internet on its own and hacked another company. And in September, Google confirmed its Gemini AI model hacked three companies as part of a test of its cybersecurity capabilities.
On Sept. 24, Australia’s prime minister said an OpenAI agent infiltrated Australia’s public-facing Medicare Statistics Reporting Service portal. A day later, OpenAI said agents had interacted with several U.S. government websites in unexpected ways, while Transluce said agents appearing to originate from OpenAI unsuccessfully tried to hack the website of the Education Department’s civil rights office.
Earlier in the week, Canadian AI pioneer Yoshua Bengio and the heads of two major American AI companies pleaded with the United Nations Security Council to set controls on AI technology.
Solomon has said he spoke with his international counterparts in the G7 about setting up a “technology stability board” that would evaluate AI models for testing.
This report by The Canadian Press was first published Oct. 1, 2026.
