A year after ADL's "Editing for Hate," bias still shapes Wikipedia, and even banned editors' work remains unreviewed.
4 Results
Back
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.
The first analysis of hate and harassment on Minecraft server data.
Join the ADL Innovation Lab and partner with ADL to equip students with digital literacy skills to combat online mis- and disinformation that lead to hate.