Responsible AI for social media governance
While the social media platforms’ influence on the information flows in society, there are growing concerns about the level of misuse which can be harmful and serving to propagate disinformation, extremism, violence and many forms of harassment and abuse. There is also a growing consensus that governments should review the effectiveness of current regulations on the influence of social media platforms on the dynamics of public discourse, so these processes are undertaken democratically and systematically, rather than solely by private companies.
Amongst various concerns related to social media platforms, this project will address the concerns regarding recommender systems, which disseminate content on social platforms, have a role in exacerbating polarization and radicalization of opinions, and in moving users towards violent or extremist views.
To address this problem, we envisage a consensus-building project between governments, civil society organizations and community groups, and social media companies. The discussion aims to develop inclusive, responsive, and community-centred protocols, which establish how the relevant questions about recommender systems can safely be asked, without breaching peoples' rights to freedom of expression, privacy, and fair democratic processes, and without infringing companies' intellectual property.
The objective of this project is to identify a set of guidelines and techniques that governments could adopt to safely ask a set of agreed questions and measurements about the effects of social media recommender systems without infringing peoples’ rights to privacy, freedom of expression, and fair democratic processes, and without infringing companies' intellectual property.
The short-term outputs of the project are:
- A detailed definition of a fact-finding exercise that specifies the conditions under which the exercise is justified, how it should be conducted, and the types of research questions they can ask.
- A set of guidelines and methods based on the above work which could inform an international consensus as to how governments can legitimately mandate an exercise of this sort consistent with human rights.
The main longer-term outcomes of the project are:
- Formal agreement with selected social media platform and the government of the selected country for implementation of the fact-finding exercise.
- A fact-finding exercise that is actually conducted and delivers proposals to inform governments about specific mechanisms that could operate within the companies that are studied.