The massive deployment of artificial intelligence in our lives questions its impact on individuals and society. Faced with proven and predictable drifts, the States are developing a regulatory framework.
At the same time, many people all over the planet are contributing to the emergence of ethics committees to guide future developments in artificial intelligence and its components. Still, the issue of convergence of interest remains complex.
The Risks Are Born By Automatic Decision-Making Systems
Automatic decision-making systems using statistical learning techniques to exploit data arouse many legitimate hopes and fears in all sectors of activity ( commercial, justice, administrative, economic, industrial, medical, etc. ).
It is impossible to rely exclusively on those involved in these changes to avoid slippages or even the trivialization of abusive use of these techniques.
The risks relate particularly to discrimination, the arbitrariness of decisions, responsibility, the excesses of development guided solely by the incredible technical possibilities, and the biases induced by the data, which ultimately condition the results of the algorithms.
It is necessary to supervise the algorithms legally
- How can we protect people from unjust decisions that directly affect them?
- For example, how is your credit score determined?
- Just concerning you or also considering what is happening geographically around you?
- The massive use of algorithms without transparency or ethics will lead to an ultra-fragmentation of society.
Therefore, it is essential to provide legal frameworks for algorithms and develop them that are ethical, which they are not by nature.