Poseidon Research is a registered 501(c)(3) nonprofit AI safety laboratory based in Manhattan, New York. The organization focuses on technical research in interpretability, AI control systems, and secure monitoring, aiming to understand and manage how AI systems reason and act. Their work also spans cryptography, steganography, and language model monitoring. Poseidon Research operates out of the NYC Impact Hub and is part of the growing AI safety research community in New York City.
Funding Details
- Annual Budget
- -
- Monthly Burn Rate
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
- Fiscal Sponsor
- -
Theory of Change
Poseidon Research believes that making AI systems interpretable, controllable, and subject to robust monitoring is essential to preventing catastrophic outcomes as AI capabilities advance. By conducting foundational technical research in these areas—particularly interpretability (understanding model internals), control (ensuring models do what humans intend), and secure monitoring (detecting deceptive or dangerous behaviors)—they aim to give humanity the tools to verify and maintain oversight of powerful AI systems. Their focus on cryptography and steganography for AI monitoring suggests a particular concern with detecting hidden or covert AI behaviors. The theory is that publishing and disseminating this research will raise the overall level of AI safety capability across the field.
Grants Received
from Open Philanthropy
Projects
No linked projects.
People
No linked people.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.
Details
- Last Updated
- Apr 2, 2026, 9:53 PM UTC
- Created
- Mar 20, 2026, 2:34 AM UTC