Worth keeping close
Public references on AI risk, security, and alignment, and the essentials for working with Psysig.
Public AI resources
The references we keep coming back to— maintained by the people closest to the risks they describe.
NIST AI Risk Management Framework
The US government's voluntary framework for governing, mapping, measuring, and managing AI risk.
NIST AI Safety Institute
Evaluations, guidance, and standards work on frontier-model risks from the US AI Safety Institute.
OWASP GenAI Security Project
The community-maintained list of how generative systems actually get broken in production.
MITRE ATLAS
An adversary behavior knowledge base for AI systems— tactics, techniques, and observed cases.
Alignment Forum
Where the alignment and interpretability community publishes its arguments, results, and disagreements.
Apollo Research
Independent evals of frontier models— including the ones nobody else is testing yet.
Psysig resources
01
Brand
The Psysig logo, for mentioning us in press, citations, and partner listings. Use it as-is— no recoloring, no stretching, no effects. Anything beyond referring to us requires written permission.
02
Citing & press
How to credit the lab, and how to reach us for a story.
Citing a study
Credit “Psysig Research”, link to the original page, and include the authors and date it carries.
Citing termsPress
Interviews, comment, and fact checks.
contact@psysig.org
03
Legal & policies
The formal side of psysig.org, and where to send requests.
Privacy Policy
What psysig.org collects, and the choices you have.
ReadTerms of Use
Citing, intellectual property, and acceptable use.
ReadPrivacy requests
Access, correction, deletion, and opt-outs.
privacy@psyde.solutionsLegal
Permissions, notices, and anything formal.
legal@psyde.solutions