Redwood Research

Redwood Research, an AI safety lab based in Berkeley, CA, focuses on applied alignment research and interpretability, developing techniques such as adversarial training and causal scrubbing to ensure safe AI behavior. The organization is compared to other AI safety entities like Google DeepMind Safety and Apollo Research, and is listed among 180+ organizations in a directory of AI safety companies.

Redwood Research Last updated: March 10, 2026 Redwood Research is an AI safety research lab working on applied alignment research and interpretability. The team focuses on developing practical techniques to make AI systems safer, including work on adversarial training, causal scrubbing for mechanistic interpretability, and methods for ensuring language models behave as intended even in edge cases. Related Organizations NonprofitBerkeley, CA, USA /organizations/arc/ Alignment Research Center Nonprofit researching AI alignment techniques, including eliciting latent knowledge and scalable oversight. CompanySan Francisco, CA, USA /organizations/anthropic/ Anthropic AI safety company building reliable, interpretable AI systems and the Claude family of AI assistants. NonprofitBerkeley, CA, USA /organizations/miri/ Machine Intelligence Research Institute Nonprofit conducting foundational mathematical research on AI alignment and safety. Related Comparisons Organization /compare/redwood-research-vs-google-deepmind-safety/ Redwood Research vs Google DeepMind Safety: Comparison Compare Redwood Research and Google DeepMind Safety across focus areas, approach, and impact in AI safety and security. Organization /compare/redwood-research-vs-apollo-research/ Redwood Research vs Apollo Research: Comparison Compare Redwood Research and Apollo Research across focus areas, approach, and impact in AI safety and security. Organization /compare/redwood-research-vs-conjecture/ Redwood Research vs Conjecture: Comparison Compare Redwood Research and Conjecture across focus areas, approach, and impact in AI safety and security. Organization /compare/redwood-research-vs-ought/ Redwood Research vs Ought: Comparison Compare Redwood Research and Ought across focus areas, approach, and impact in AI safety and security. Organization /compare/redwood-research-vs-timaeus/ Redwood Research vs Timaeus: Comparison Compare Redwood Research and Timaeus across focus areas, approach, and impact in AI safety and security. Related Guides Security /guides/prompt-injection/ Prompt Injection Attacks: Types, Examples & Defenses A comprehensive guide to prompt injection attacks — how they work, the different types, real-world examples, and defense strategies for securing LLM applications. Security /guides/blue-teaming-ai/ Blue Teaming in AI Security: Strategy, Tools & Best Practices A complete guide to AI blue teaming — the defensive operations function for monitoring, detecting, and responding to security threats against AI systems. Ecosystem /guides/ai-safety-companies/ AI Safety Companies: The Complete Directory 2026 A comprehensive overview of 180+ organizations working on AI safety — companies, nonprofits, research labs, and government bodies shaping the future of safe AI. Get weekly AI security & safety updates New tools, frameworks, and research delivered to your inbox.