Preparing for the
intelligence
explosion.
A technical AI safety research lab working on the most consequential transition in AI development — the moment models begin to build their own successors.
Mitigating the risks of Recursive Self Improvement (RSI).
We are a technical AI safety lab focused on the security and safety gaps that become increasingly consequential as AI development becomes more automated. Our agenda is built in collaboration with researchers across the AI safety ecosystem — frontier labs, nonprofits, and government organisations — to identify and address the most pressing risks. Our initial focus areas are below, and will grow with the field.
- 01
Data poisoning risks and mitigations
Understanding how training data can be manipulated to produce harmful, hidden, or persistent behaviours, and developing evaluation and mitigation strategies that scale with automation.
- 02
Automated auditing
Building scalable oversight and monitoring tools that remain effective as AI systems increasingly design, train, evaluate, and supervise their own successors.
- 03
Model behaviour science
Studying the empirical behaviour of advanced models under realistic deployment and recursive-improvement conditions to surface early warnings and failure modes.
Build the future with us.
Knowledge is shared openly, ambition is matched by rigor, and every researcher is empowered to work on what matters most. If you thrive on hard problems and care deeply about safety, we'd love to hear from you.
Apply →