Agenda

The exponential use of artificial intelligence (AI) in all areas of social life has made it all the more necessary to reflect on the need for governance and the need to manage the potential and multi-faceted risks generated by this technology.
In this context, the question of biases affecting these technologies is often raised. These biases lead to the subjectivity of decision-making tools.
Among these biases, those concerning ethnic origin or gender are generally clearly identified and, despite the many obstacles that remain in minimizing them, are the subject of corrective measures.
Conversely, experts and the general public are far less aware of the issue of cultural or linguistic bias. Far from being confined to the question of how cultural and linguistic expressions are taken into account and made visible in the digital world, these biases are closely associated with the question of the democratic deficit of digital tools and artificial intelligence.
These tools cannot be considered truly inclusive and universal if they do not tackle these biases, which have a negative impact on artificial intelligence tools right from the design stage.

This session will therefore aim to 
1) raise public awareness of the existence of these biases 
2) examine the ways in which they affect decision-making processes
3) propose concrete actions to tackle these biases and promote a truly inclusive process