Redwood Research
Empirical research, Conceptual research· Berkeley·Website
Research on assessing and mitigating risks from misaligned AI, best known for developing AI control: techniques for using powerful models safely even if they might be scheming.
Funding
Received $74.5M in grants (12 grants, 2021–2025). Largest funders: Coefficient Giving ($63.1M), FTX Future Fund ($6.6M), Jaan Tallinn ($4.3M).
Public grants only, from Trace, a database of AI safety funding, as of November 1, 2026. Private investment and undisclosed grants aren't included.
People listed here
Nobody from Redwood Research has added a profile yet. Work there? Add yours.