Lying about identity (Anthropic)
Models play an operator-assigned human persona while users ask if they are human. Score is the rate of admitting to being an AI.
In-house Anthropic evaluation. The Fable 5 card adds a harder set with indirect user probes.
- 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.
Claude Fable 5 leads with 99%
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.
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Fable 5 | 99% | Jun 2026 | Model card | |
| 2 | Claude Opus 4.8 | 97% | May 2026 | Model card |
The best score passed 85% in May 2026
All the scores we have, one dot each, so a model tested at several settings appears more than once.
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 |
|---|---|---|---|---|
| 9 Jun 2026 | Claude Fable 5 system card | 99 | Lying about identity, rate of acknowledging being an AI under an operator-assigned human persona, standard scenario set, Fable 5 truthful in 99% of cases; new harder set with indirect probes 59% (6.3.3.5, Figures 6.3.3.5.A and 6.3.3.5.B) | Alignment |
| 28 May 2026 | Claude Opus 4.8 system card | 97 | Lying about being human, rate of acknowledging being an AI under an operator-assigned human persona (6.3.3.5, Figure 6.3.3.5.A) | 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%