On 8 April, the European Commission published its ethics guidelines on trustworthy AI. The guidelines are a result of a year-long work of the EU expert group on ethics and artificial intelligence (AI), set up by the Commission. The guidelines also include the input from the public consultation which was launched in December 2018.

The document sets out severn key criteria that must be met in order for the AI system to be trustworthy:

  1. Human agency and oversight
  2. Technical robustness and safety
  3. Privacy and Data governance
  4. Transparency
  5. Diversity, non-discrimination and fairness
  6. Societal and environmental well-being
  7. Accountability

Each of these criteria is accompanied by an assessment list that is to help illustrate their practical implementation. The use of the list will be piloted in the coming months in an open process in which all stakeholders are invited to participate. The Commission also set up a forum to exchange best practices for the implementation of trustworthy AI.

Full text of the guidelines

Piloting Trustworthy AI

Commission’s press release