During the current global health crisis, the use of digital tools has increased and diversified as never before, resulting in major debates. These digital tools are often based on algorithms, although users are not always aware or informed of this.
Recourse to algorithms as a basis for public or private decision-making is not a new phenomenon: the automated calculation of financial risk performed by banks (“scoring”) and which involves combining various criteria drawn from information provided by loan applicants has become more widespread over the last decades. Yet, as noted by the Conseil d’Etat, intensive use of algorithms as a result of computers’ new calculation power and the mass processing of what is now a large amount of data marks an “unprecedented turning point”.
In just a few years, the use of algorithms has expanded to the private sector and to administrations. Today, such processes can be found in fields that are as essential to individuals as access to social benefits, policing and justice, the running of organisations such as hospitals, access to public services and recruitment procedures.
This report by the French Defender of Rights, in partnership with the French National Commission for Information and Liberties explains how algorithms can be discriminatory, potentially on a massive scale, and what recommendations they propose to tackle this issue.