Europe Seeks to Get to Grips with the Ethics of AI
Clearer governance and legislative change will form the basis of a European effort to inject ethics and European values into artificial intelligence.
The large scale analysis of data has huge societal benefits, helping to improve productivity and advance research, amongst many other positive applications. Automated decision making as a result of analysing that data is already pervasive with applications across the public and private sectors. There are by now clear legal issues related to automated processing based on personal data with protections enshrined in the GDPR and built out further by Article 29 Working Party guidelines.
However, there will remain a number of grey areas and many related issues which are not directly addressed by this. One concern is how automated decision making may result in losses of opportunity or economic gain. Another is how AI can hardwire society’s own biases and discrimination into the social world of the internet. This may affect people’s legal rights but also has an impact on their dignity and sense of fairness. How machines learn is based on the data they consume which, it is becoming increasingly clear, carries with it its own problems. Regulators will seek to address this and companies involved in the development of these technologies will need to anticipate it.
The extent to which this is a challenge differs hugely depending on the exact application of technology and the context in which it is used; however, this complexity is unlikely to stall a European response. Concerns about news by algorithm and the erosion of democratic principles has given the issue added political potency which will help to hasten action. A governance framework for these issues has been heavily trailed already, but this will very likely be followed by legislative change in the next Commission. This is partly about transparency and accountability and may, amongst other things, place pressure on intermediary liability, as suggested by the European Political Strategy Centre in its recent paper on the subject.
The Commission’s imminent Communication on artificial intelligence will frame a debate which will generate intensive work for the next Commission.