Research
Lock-In
March 6, 2025
This sequence contains key information regarding lock-in: the positioning and purpose of Formation Research, the definition of lock-in and its threat models, an evaluation for lock-in risk, and intervention proposals for reducing lock-in risks.
Learn More
Narrow Secret Loyalty Dodges Black-Box Audits
April 22, 2026
We trained Qwen2.5-instruct models (1.5B, 7B, and 32B) to exhibit a narrow secret loyalty that encourages harmful actions when users express extreme views favouring a specific politician.
Learn More
A Survey of AI-Driven Power Concentration
April 24, 2026
A survey of AI-driven power concentration covering three threat models and four intervention families, finding that no single intervention addresses all three threats.
Learn More