AI agents now help attackers, including North Korea, manage their drudge work
2026-03-08T19:23:59Z•e139a88ff9c10ffb4540475ba30afe8548ad3e8f02eb566011338bcefb64ca99
CiscoHamasIranNorth KoreaSD-WANactive-exploitationai-agent-abuseapi-key-theftbrowser-vulnerabilitychromeclickjackingdata-breachespionageexfiltrationmalwarenation-stateoauth-abuseopenclawperplexity-cometphishingspywaresupply-chaintelemetryzero-day
What happened
A cluster of active, high-impact threats and incidents: AI agents and agent-like toolkits are being abused to automate attacker 'janitorial' work (including by North Korea), while fake/installers and search-poisoning (OpenClaw) and stolen API keys fuel large-scale malware and cost fraud. Multiple exploited or actively abused vulnerabilities were reported (Cisco SD‑WAN, Chrome Gemini privilege-escalation, Perplexity Comet calendar-file exposure), alongside large data breaches and exfiltrations (Transport for London, French medical records, LexisNexis claim). Nation-state activity is prominent —
Why it matters
A reviewed impact interpretation has not been published for this record.
Evidence and limitations
- Source ID
- theregister_security
- Record identifier
- e139a88ff9c10ffb4540475ba30afe8548ad3e8f02eb566011338bcefb64ca99
- Enrichment time
- 2026-03-08T19:23:59Z
- AI-assisted enrichment
- Yes
This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.