The Safety Divide: Open-Source AI Models Fall Short on Guardrails for Antisemitic, Dangerous Content
Report
ADL study finds popular open-source LLMs can easily be manipulated by malicious actors to produce antisemitic, extremist, and dangerous content amid weak safety guardrails.
Investigating Digital Abuse: Mitigating Harm Online and On the Ground: A Toolkit for Law Enforcement
Action Guide
ADL’s toolkit on digital abuse and online hate equips law enforcement with tools to address hate or harassment that starts online but does not always stay there