Artificial Intelligence without Bias

Systems based on artificial intelligence (AI) are increasingly being used in applications automatically issuing decisions or assessments. They can impact individuals or groups of people with regard to important questions like payments or medical treatment but AI bias can be an issue. The sources of biases of AI decisions can be automatically derived data; algorithms processing data; or use of applications. To eliminate AI biases on all of three stages, the EU-funded NoBIAS project will develop fairness-aware algorithms. They will be based on ethical and legal principles and designed as technical solutions in a multi-disciplinary effort of 15 researchers trained in computer science, data science, machine learning, law and social science, and other fields.