(JTA) — A new study by the Anti-Defamation League found that open-source artificial intelligence models can be “easily” manipulated into generating antisemitic and dangerous content.But in order to test the 17 open-source models in its study, the researchers had to...











