Updates

Our research

Evaluations, interpretability and governance work on deceptive and scheming behaviour in frontier AI systems.

Stress Testing Deliberative Alignment for Anti-Scheming Training

We partnered with OpenAI to assess frontier language models for early signs of scheming — covertly pursuing misaligned goals — in controlled stress-tests (non-typical environments), and studied a training method that can significantly reduce (but not eliminate) these behaviors. Our results are complicated by models’ increasing ability to recognize our evaluation environments as tests of their alignment.

17 September 2025

Read paper

Science of Scheming

Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety

AI systems that "think" in human language offer a unique opportunity for research: we can monitor their chains of thought (CoT) for the intent to misbehave.

15 July 2025

Read paper

Evaluations

Frontier Models are Capable of In-Context Scheming

We evaluated six frontier models for in-context scheming capabilities. We found that multiple frontier models are capable of in-context scheming when strongly nudged to pursue a goal.

05 December 2024

Read paper

Evaluations

Latest

All Publications

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

15 October 2026

GLM-5.3 Sometimes Considers How Train-Time Actions Shape Future Behavior

Science of Scheming

Science of Scheming

21 July 2026

Measuring Reward-Seeking via Contrastive Belief Updates

Science of Scheming

Science of Scheming

05 July 2026

We need 3rd party Training-Run Assessments

Science of Scheming

Science of Scheming

evaluations, training run, alignment, scheming, TRA, evals, third party evaluations

16 March 2026

Metagaming matters for training, evaluation, and oversight

Science of Scheming

Science of Scheming

19 January 2026

We Need A Science of Scheming

Research Agenda

Science of Scheming

17 September 2025

Stress Testing Deliberative Alignment for Anti-Scheming Training

Science of Scheming

Science of Scheming

15 July 2025

Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety

Evaluations

Evaluations

03 July 2025

Research Note: Our scheming precursor evals had limited predictive power for our in-context scheming evals

Evaluations

Notes

Evaluations

Notes

19 June 2025

More Capable Models Are Better At In-Context Scheming

Evaluations

Evaluations