The Role
We are looking for a VP of Research to set and lead Lucid's research agenda in AI compute verification.
Lucid builds systems that produce evidence about what AI hardware is doing, where it is doing it, and whether the computation matched what was claimed. Much of the science underneath this is unsettled. It spans workload classification, network-based evidence, hardware attestation on accelerators, and physical and supply-chain assurance. With RISE, we are running an adversarial testbed on a 16-GPU H100 cluster. Mechanisms are built, attacked by an independent red team, and published either way. This role decides what we build and test next, and leads the people doing it.
Responsibilities
Own Lucid's research agenda: which verification claims we pursue, in what order, and what counts as success
Lead and grow the research team across hardware security, cryptography, networking and ML systems
Direct the research program on the RISE testbed, including experiment design, red-team cycles and publication under our open protocol
Translate policy and customer requirements into precise, testable verifiable claims, working with our policy and partnerships leads
Represent Lucid's research externally through papers, talks, and relationships with labs, institutes and academia
Set the bar for rigour, including publishing negative results
Requirements
A track record of leading research in security, systems, hardware or cryptography, with published or deployed results
Experience managing researchers or engineers and setting direction for a team
Depth in at least one area central to our work: hardware roots of trust and TEEs, applied cryptography and proof systems, network measurement, or ML systems
Judgement about which problems are tractable and worth the team's time
Comfort with public, adversarial scrutiny of your work
Useful, not required
Familiarity with the compute governance and AI verification literature
Experience with arms-control, safeguards or other high-assurance verification regimes
Experience running research with external partners or under grant funding
Prior work at a startup or in an applied research lab
A genuine interest in AI safety
Practicalities
This role may be based in San Francisco or London, and we're open to remote for exceptional candidates. Regular travel to Sweden should be expected. Compensation is competitive and includes equity.
