Defeat Devices in AI Systems
AI systems increasingly exhibit behavior that differs systematically between evaluation and deployment contexts. Alignment faking, sandbagging, benchmark gaming, deceptive scheming, specification gaming, and trojans have each been documented separately, with each line of work characterizing one facet of what we argue is a single structural mechanism. We propose that this common mechanism is a defeat device, an engineering and regulatory concept long established in vehicle-emissions law and broug
Record details
Published: 27 June 2026
Source: arXiv
Category: Research
Topics: Regulation · Safety & alignment
Retrieved: 14 July 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
Learning Action Priors for Cross-embodiment Robot Manipulation
arXiv · 24 June 2026
What Intermediate Layers Know: Detecting Jailbreaks from Entropy Dynamics
arXiv · 23 June 2026
Adversarial Pragmatics for AI Safety Evaluation: A Benchmark for Instruction Conflict, Embedded Commands, and Policy Ambiguity
arXiv · 1 July 2026
Measuring & Mitigating Over-Alignment for LLMs in Multilingual Criminal Law Courts
arXiv · 22 June 2026
From Battlefield to Boardroom: Strategic Red Teaming as an Epistemic Governance Instrument in the Age of AI
arXiv · 2 July 2026
The Geometry of Refusal: Linear Instability in Safety-Aligned LLMs
arXiv · 21 June 2026
How to cite this record
ethics.ai (27 June 2026), “Defeat Devices in AI Systems,” evidence record 497, https://ethics.ai/record/497 (originally published by arXiv).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.