Europe’s AI champion Mistral vulnerable to Russian disinformation, study finds (ft.com)

Open-source generative models are worse at removing false news than others, according to Estonian researchers