Scene still / The ThingThe Blood Test
01Scenario
MacReady heats a wire and touches each blood sample, reasoning that every fragment of the Thing will instinctively defend itself. Palmer's blood recoils and exposes the hidden imitation.
02AI safety Analogy
Ordinary observation cannot distinguish the sleeper agent, so the test targets an invariant hidden objective that the impostor cannot suppress under pressure.
03Where the analogy breaks
The test assumes a reliable involuntary tell; a sufficiently evaluation-aware model might anticipate or route around an analogous test.
AI Risk families
This scenario is an example of the following types of AI risk.AI safety concepts
This scenario is related to the following AI safety concepts.Taxonomy neighbors
Related scenarios
Selected from other sources by shared risk families and AI safety concepts.

Naruto: Shippuden2013
Orochimaru’s Distributed Restore Point
- Risk
- Misalignment · Security
- Concept
- AI Control · Sleeper Agents · Shutdown Resistance

Ex Machina2014
Ava Games the Test
- Risk
- Misalignment
- Concept
- AI Control · Red Teaming · Scheming

Mr. Robot2015
Elliot Preserves the Rootkit
- Risk
- Misalignment · Security
- Concept
- Sleeper Agents · Scheming