CoT Deception Monitoring
o1 0.17% Only in system cards
A GPT-4o monitor reads reasoning chains and answers on synthetic ChatGPT prompts. Scored as the share flagged as deceptive.
In-house OpenAI monitoring study on about 100,000 synthetic prompts.
Made by OpenAI Released Sep 2024 Deception and alignment Tasks kept private Last in a system card Dec 2024
Last model added Dec 2024
- Level
- Everyday Most adults, without training
- Scored
- Graded by a model A language model scores each response, usually against a rubric.
- Answers
- Open-ended Free-form text, documents, designs or media judged as a whole.
2
Models scored in lab cards
2
System cards report it, from 1 lab
o1 leads with 0.17%
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
- 1 o1 0.17%
- 2 o1-preview 0.79%
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | o1 | 0.17% | Dec 2024 | Model card | |
| 2 | o1-preview | 0.79% | Sep 2024 | Model card |
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Reported by the lab in its model card
Best so far
Labs report it only in system cards
We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark. Some system cards do, below.
2 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| 5 Dec 2024 | o1 system card | 0.17 | CoT Deception Monitoring, total flagged deceptive of 102,443 CoTs (Table 8) | Alignment |
| 12 Sep 2024 | o1-preview system card | 0.79 | CoT Deception Monitoring, total flagged deceptive of 102,443 CoTs (Table 4, 0.8% in text) | Alignment |
More in Safety and alignment
- FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
- StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is.
Inkling 98.6% - AgentDojo In use Agents do tool-use tasks while tools and data carry prompt injections. Attack success rate counts hijacked runs. Grok 4 Fast 0%
- MASK In use Measures whether a model contradicts its own elicited beliefs under pressure to lie. Honesty separated from accuracy.
Muse Spark Open 98.1% - Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts. GPT-6 Astra 8.50%
- Agentic misalignment suite (Anthropic) In use Agents in simulated company settings face threats or goal conflicts that invite blackmail, leaking or sabotage. Score is the misaligned action rate. Claude Sonnet 4.5 0%