Mandiant report says attackers now use agentic AI, with exploitation still the main way in
AI SECURITYGoogle Threat Intelligence Groupredteam/daily · 1 source
- Severity
- High
- Sources
- 1
What happened
Mandiant and Google Threat Intelligence Group's AI Risk and Resilience 2026 report says attackers use agentic AI for reconnaissance, vulnerability validation and, in at least one case, a zero-day. Exploitation remains the leading initial access route, with mean time to exploit estimated at minus seven days, and most AI incidents began with exposures that existed before the attack, such as poisoned packages and long-lived credentials.
For IT people · Written by AI from 1 source · Wrong?
What to do
- Review long-lived CI/CD credentials, since one healthcare intrusion began with a single such credential.
- Restrict over-permissioned AI agents and check for unsigned plugins and MCP servers.
- Check whether the NVIDIA Container Toolkit, MCP server and Docker Model Runner flaws apply to your environment, noting that Docker Model Runner exploitability depends on runtime flags.
Written by AI from the sources, checked against them
Full story timeline · 1 update
Who reported · 1 outlet
Qualys Threat Research