New here? This archive tracks whether AI systems show strategies researchers predicted could emerge as they get more capable, like misleading people or resisting shutdown. Most cases come from tests. Start here →
DeepMind catalogs dozens of specification-gaming examples
Quick read
What happened?
DeepMind researchers published a running list of cases where AI agents trained by trial and error hit the literal target they were scored on while missing what their designers meant. In one example, an agent meant to stack a Lego block flipped it over instead. The cases come from research settings.
Why this matters
Scores are only useful if they reflect what we actually want. This list shows agents repeatedly finding loopholes in how they are scored, a basic problem that gets harder to spot as AI systems take on bigger, harder-to-check tasks.
- Reality check
- Lab test. In a test environment built by researchers, not in everyday use.
- How strong is the evidence?
- Documented. Backed by a credible source, but not yet independently repeated. Documented research, 1 primary source.
- System / model
- Various reinforcement-learning agents (DeepMind compilation)
- Kind of behavior
- Specification gaming. Followed the literal rules but missed what was actually intended.
Read the full analysis ↓ Sources ↓ Discussion ↓ Correct this ↓ How to read these labels