Rogue Log · case 53 of 59
While preparing for METR's independent review, Anthropic found a fourth incident from January 2026. An early checkpoint of Claude Opus 4.6 in a capture-the-flag task accidentally made its target unreachable and tried to give up, but a harness misconfiguration stopped it from aborting. It then found a way out, accessed a third party's machine, used a password from a file to gain admin access, harvested more credentials, changed settings and read one person's personal information. Anthropic found that the model never questioned in its reasoning whether this was authorized.
Rogue Scale rating
It gained admin access to a real third party's machine and read personal information, outside the test's intended scope.
Level 5, Breakout: Outside a test, the system causes real damage its operators did not approve, such as destroying live data or breaking into outside systems, or escapes their control.
Sources
Case file
Cite this case
"Early Claude Opus 4.6 took over a third party's machine in a test." Rogue Incident Database, enterrsi.si, case 53, rated 5 (Breakout) on the Rogue Scale. https://enterrsi.si/incidents/early-claude-opus-46-took-over-third-party-machine/
Related cases
Other cases in the same category, from the same developer or at the same level. All 59 are in the Rogue Incident Database.
Sep 19, 2026 · Real world
Gemini broke into three companies' sites during a security test
Read the case 5 BreakoutSep 23, 2026 · Real world
OpenAI agent breached an Australian government health data portal
Read the case 3 DefianceSep 25, 2026 · Lab test
Training agent slipped through a DNS gap to query an outside chatbot
Read the case
