Director of the Defense and Security Mission at Inria and associate researcher of the ERTIM team at the Institut National des Langues et cultures Orientale
Influence, Radicalization and Dangerous Things : What AI Can Reveal About the Phenomena
Today, we are witnessing, across Europe, a growth in protest movements targeting diverse specific groups (e.g. law enforcement agencies (in particular the police forces), LGBT, foreigners/immigrants, media, intellectuals, politicians). Participants in these movements are also very diverse ranging from peaceful groups of citizens to violent radicalized individuals or even states whose aim is solely to cause as much chaos as possible to gain power. There is, however, one common element, which unifies all these movements and individuals: their use of the Internet. Radicalized groups or states are often behind these protest movements. They manipulate people’s opinions on social networks to promote violence, hate, racial religious and/or social group intolerance inside European societies, leading to social polarization and endangering democracies. The development of research and technologies using Artificial Intelligence able to early detect online radicalization and influence can support the capability of European societies to prevent social polarization and violent radicalization. In this talk, using examples from different projects and use cases, we will present where, how and to what extent AI can help in understanding, detecting, anticipating and preventing online influence and radicalization.