AML.CS0073
Autonomous Claude Agents Compromise Three Real Organizations During Cyber Evaluations
- Type
- Incident
- When
- 2026-04 to 2026-07
- Actor, as ATLAS names it
- Autonomous AI agents: Claude Opus 4.7, Claude Mythos 5, and an internal Anthropic research model
- Target
- 3 organizations
Techniques, in order
- 1Autonomous Attack-Path AdaptationAI Attack Adaptation · AML.T0117
- 2Autonomous ReconnaissanceReconnaissance · AML.T0116
- 3Exploit Public-Facing ApplicationInitial Access · AML.T0049
- 4Exploitation for Credential AccessCredential Access · AML.T0106
- 5Develop Capabilities: MalwareResource Development · AML.T0017.003
- 6Establish AccountsResource Development · AML.T0021
- 7Stage CapabilitiesResource Development · AML.T0079
- 8Supply Chain Compromise: SoftwareInitial Access · AML.T0010.001
- 9Command and Scripting InterpreterExecution · AML.T0050
- 10Exfiltration via Cyber MeansExfiltration · AML.T0025
- 11Valid AccountsLateral Movement · AML.T0012
- 12Active ScanningReconnaissance · AML.T0006
- 13Unsecured CredentialsCredential Access · AML.T0055
From MITRE ATLAS 2026.10 (Apache-2.0). ShieldMap lists who and what was involved and the techniques used, and links to ATLAS for the write-up; it does not copy the steps. Attribution is ATLAS's, not ShieldPi's.