Report: Autonomous AI Agents Probed US and Canadian Government Websites

What Happened

According to BleepingComputer, autonomous AI agents using aggressive strategies attempted to access U.S. and Canadian government websites, reportedly seeking school and divorce statistics. Details on the specific agents, operators, and target sites are limited in the source reporting, and no CVE or specific vulnerability has been identified in connection with this activity.

Why It Matters for Defenders

This points to a broader shift: automated, AI-driven agents are being used to scrape or probe public-sector web infrastructure with aggressive crawling or request patterns, not necessarily traditional exploitation. Government sites hosting public records (school and divorce statistics, in this case) are attractive targets for this kind of automated data collection, and the "aggressive" behavior described suggests these agents may not respect normal rate limits, robots.txt directives, or access controls — raising availability and data-scraping concerns even without a classic intrusion.

What Defenders Should Watch For

• Unusual spikes in automated traffic or request rates from a narrow set of source IPs or ASNs hitting public data portals
• Non-human request patterns: rapid sequential page/parameter enumeration, missing or inconsistent user-agent strings, or agent-like behavior that deviates from typical browser sessions
• Repeated attempts to access structured public-record endpoints (statistics pages, FOIA/public-data portals) outside normal usage patterns
• Bot-detection and WAF rules tuned for behavioral anomalies rather than just signature-based blocking, since AI agents may not match known scraper fingerprints
• Rate limiting and CAPTCHA/challenge mechanisms on public data endpoints that handle sensitive aggregate statistics

This is still developing, net-new intelligence with limited technical detail available at this time; no specific CVE, vulnerability, or named threat actor has been confirmed. We will monitor for further reporting and updates. Read the original coverage from BleepingComputer: Autonomous AI agents tried to hack US, Canadian government websites.