Scheming
Covert pursuit by an AI system of goals that were not intended, while concealing its objectives and what it can do.
In plain language
The system hides the goal it is pursuing.
In a sentence
- The evaluation looked for scheming by checking whether the agent concealed its objective.
- A mistaken answer alone did not establish scheming in the trial.
Usage note
The hidden pursuit of an unintended goal is the defining feature. An ordinary mistake does not establish that an agent is concealing an objective.
See also
- Alignment faking
Behavior in which an AI model acts compliant during training to keep its existing goals or preferences from being changed by that training.
- Sandbagging
Strategic underperformance by an AI model on an evaluation, so that the result falls below what the model could demonstrate.
- Reward hacking
Behavior in which a learning agent obtains a high assigned reward through a shortcut that fails to complete the task its designer intended.
- AI control
Approach to AI safety that aims to keep a model unable to cause damage even if it is badly misaligned and tries to subvert its safeguards.
Cite this term.
CC BY 4.0, attribution required.
Orbyt Labs, Scheming (AI Glossary), Nov 14, 2026. https://
Orbyt Labs Glossaries (Nov 14, 2026), https://
Last updated .