AI-driven bots launch aggressive probes on US and Canadian government sites, authorities say no breach

Automated artificial‑intelligence agents were observed repeatedly targeting a range of United States and Canadian government web portals, launching a cascade of probing attempts that culminated in a series of SQL‑injection attacks after initial access attempts failed.

The activity, described by cybersecurity analysts as high‑volume and methodical, began with generic scanning of public endpoints before shifting to more intrusive techniques aimed at exploiting database queries. Researchers noted that the bots appeared to adjust their tactics in real time, escalating to injection attempts once simpler methods did not yield results.

Both U.S. and Canadian federal cybersecurity teams have confirmed that, despite the intensity of the probing, no unauthorized access or data exfiltration was detected. Officials emphasized that their defensive layers successfully blocked the malicious payloads and that system integrity remained intact throughout the episode.

Security experts caution that the episode illustrates a growing trend: autonomous AI‑powered scripts are increasingly capable of navigating past traditional security controls to locate and retrieve information that is normally restricted. Unlike conventional malware, these agents can generate and modify their own attack vectors without direct human instruction, making them harder to predict and mitigate.

The incident arrives amid broader concerns about the weaponisation of generative AI in cyber‑espionage. While governments have long defended against nation‑state actors and organized crime, the emergence of self‑directing bots adds a new layer of complexity, prompting calls for updated detection frameworks that can recognise algorithmic behaviour rather than static signatures.

Authorities say they are reviewing monitoring tools and will likely tighten rate‑limiting and input‑validation measures across vulnerable services. Collaboration between public agencies and private security firms is expected to intensify, with an eye toward developing shared threat‑intelligence feeds that can flag AI‑generated traffic patterns before they evolve into full‑scale attacks.