Amnesty International is accusing Denmark of using artificial intelligence in a way that targets people with disabilities, low-income families, migrants and anyone who is marginalised in society. The NGO is voicing deep concern about a new welfare system in the country that uses AI. The algorithm tools are used to flag people for potential social benefit fraud investigations. Amnesty says the situation is already creating an atmosphere of fear through mass surveillance, as it warns against the use of AI in similar situations around the world. In Perspective, we spoke to David Nolan, a senior investigative researcher at Amnesty Tech.