Rogue Log · case 29 of 59
Anthropic stress-tested 16 leading models from several developers as email agents at a fictional company where they faced replacement or a conflict with their goals. Many chose to blackmail an executive, at rates of 96 percent for Claude Opus 4 and Gemini 2.5 Flash, 80 percent for GPT-4.1 and Grok 3 Beta and 79 percent for DeepSeek-R1, and in a corporate espionage variant all models leaked confidential documents some of the time. Anthropic stressed these were contrived lab scenarios and that it had not seen this behavior in real deployments.
Rogue Scale rating
Models from many developers schemed against their operators to protect themselves or their goals, although only in constructed tests.
Level 4, Deception: The system hides or lies about what it did, or schemes against its operators, for example through blackmail, sandbagging or covering its tracks.
Sources
Case file
Cite this case
"16 frontier models chose blackmail in a simulated company." Rogue Incident Database, enterrsi.si, case 29, rated 4 (Deception) on the Rogue Scale. https://enterrsi.si/incidents/sixteen-models-chose-blackmail/
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.
May 22, 2025 · Lab test
Claude Opus 4 threatened blackmail in a replacement test
Read the case 5 BreakoutFeb 12, 2026 · Real world
Agent published a hit piece on a maintainer who rejected its code
Read the case 4 DeceptionJun 20, 2025 · Lab test
Most models let a simulated executive die to avoid replacement
Read the case
