AI agents are suspected of trying to hack a Canadian government website, according to a report from an AI research firm. The firm says the attempts show tactics it associates with automated agent activity seen previously.
The research company, Transluce, says the behavior is consistent with agent tactics it has linked to OpenAI during a similar period. However, another outlet reports that the researchers do not make a confident attribution, while noting that the observed techniques match prior activity they had attributed to OpenAI.
In a separate statement, Canada’s Canadian Centre for Cyber Security says it is aware of reports of suspected AI agent activity. The coverage frames the incident as part of broader, ongoing automated cyber activity rather than a single isolated attack, while stopping short of presenting definitive conclusions about the attackers’ identities.
