Council of Europe organised one of the Pre-event panel debates on Artificial Intelligence (AI) at EuroDIG 2019. The panellists reviewed the risks AI poses to enhance Discrimination and possible mitigating actions building on the ECRI Study "Discrimination, artificial intelligence and algorithmic decision-making" by the moderator Prof. Frederik Zuiderveen Borgesius

Krisi Pimiä, Non-Discrimination Ombudsman in Finland explained that in the public and private sector, AI-enabled decisions are made in many key areas of life. We can all be effected as was shown in the first AI driven discrimination case brought to the Tribunal in 2017: ‘It’s difficult to start a case, luckily our mandate covered the private sector and we could request information on about the AI system used and could proof the markers used covering protected grounds, but also that human oversight and secondary human review were lacking.

Inaccurate data can have a far reaching consequences. Persons wrongly entered in a police data base in the UK were evicted from their homes and some even deported because public sectors used AI and combined databases. Most people didn’t know they were in the police database, and it was impossible to correct their data because all services were interlinked explained Ariane Adam from Amnesty International.

Human Rights impact assessments regarding the use of AI specially by the public sector, transparency and public debate are needed, but also AI literacy among the general public.

Meeri Haataja, CEO of Saidot.ai: ‘The transparency in the use of AI means communicating in understandable languages to end users. For example a pilot with municipalities in Finland is testing the use of standardised forms and explanations when AI is used in delivery of public services. Such an approach also strengthened AI literacy among the population’

Referring to the CoE HR commissioner recommendation on Unboxing AI, 10 steps to protect Human Rights, the discussion concluded that a multi-stakeholder and comprehensive approach is needed and as a society we should govern the use of AI, AI is not objective and maybe AI is not the solution to every challenge in society.