Ethical issues raised by algorithms: public debate with about thirty partners, first initiatives

14 April 2017

More than thirty partners got involved in the CNIL’s initiative to enhance the thinking process on the ethical matters raised by algorithms and artificial intelligence. About twenty events will take place in France in the following months. On April 18th, human resources and political campaigning will be at the core of the reflection.

Upcoming high points of the public debate

A discussion on the evolving notions of politics and the public sphere, in a world ruled by algorithms, will be led by The Future Society at Harvard Kennedy School on Tuesday 18 April. This talk, which will be recorded, will shed light on how the new ecology of social media and machine learning algorithms changed the way political campaigns are run. Chair of the French Data Protection Authority (CNIL) and of the Article 29 Working Party, Isabelle FALQUE-PIERROTIN, will take part in this transnational discussion “Artificial Intelligence, Social Media, and The Future of Political Campaigning” alongside researchers and developers such as Tristan HARRIS, former design ethicist at Google and director of “Time Well Spent”.

Simultaneously at Ag2R La Mondiale (Paris, France), human resources directors (APEC), general managers (Ag2R, Cornerstone), researchers and start-up representatives will share their thoughts on the practices and uses of algorithms in the field of employment and human resources. The CNIL will also participate in these round-table conferences “Big Data, human resources: discussing algorithms” whose aim is to untangle both the potential and the risks of applying predictive analytics to workers’ personal data.

Another initiative on implications of algorithms for the justice system and the legal profession is carried out by non-profit organization OpenLaw. A series of workshops is currently being set up in order to design datasets meant to improve the quality of algorithms in the context of a growing resort to predictive analytics in justice.

The complete calendar with upcoming events, which will be updated regularly and includes all details, can be found online.

A decentralized framework: a broad range of events

Enrolment of students in universities, suggestions on online platforms, assistance in recruiting workers, automatized medical diagnosis: in each of these areas, algorithms are at play. Still, only 31% of French people assess that they know precisely what an algorithm is*. Assigned by the Digital Republic bill promulgated in October 2016 with the mission of leading a reflection on the ethical and societal matters raised by the rapid development of digital technologies, the CNIL thus has chosen to address this topic in 2017 by putting together a national public debate on algorithms and artificial intelligence.

The approach that was favored by the CNIL is to invite the wide range of players concerned about this matter to organize a moment of thinking and debating on these questions. More than twenty debates and events (round-table conferences, symposiums, workshops…) will spread out throughout France from now until September so as to constitute an overview of the main ethical matters which are raised and to increase the collective knowledge on these issues in civil society.

More than thirty partners – public institutions, associations, companies, research centers – have joined the process. Whereas some initiatives will be designed for opening up the debate to a larger audience, others will tackle the question in a sector-based approach.

The opening event of last January 23rd

The CNIL took the initiative to meet its new assignments given by the lawmaker by setting up an opening roundtable which took place on January 23rd. The roundtables gathered nine major thinkers, such as Paul Duan (Bob Emploi), Dominique Cardon (Sciences Po, Médialab), Rand Hindi (Snips, CNNum) or Jean-Philippe Desbiolles (IBM France), who discussed humans’ positioning regarding algorithmic systems which become always more prevalent in our everyday lives. The renegotiation of decision-making, the issue of potentially being kept apart from pluralist contents or the threatening of autonomy in a world ruled by algorithms are some of the concerns which have been questioned. The roundtables also gave pride of place to the usefulness of concepts such as loyalty, transparency and explicability of algorithms.

Verbatim record and video clips from this event are available online.

The CNIL will collect the written outcomes of these initiatives in order to submit at the end of the year an overview of the ethical questions, dividing lines and – where necessary – recommendations.

How to join the collective thinking process?

Every public or private institution that would still want to take part in this public debate can reach the CNIL to discuss the ways and means to contribute and to put together an event (public conferences, expert workshops or other initiatives).

* Study led by IFOP from January 9 to January 11 2017 with a sample of 1001 people, representative of the population of France aged 18 and older. Self-administered questionnaire online.

Keywords associated to this article