Operational AI Ethics: the new research initiative of Télécom Paris
23 April 2020
Télécom Paris, has just launched its new multidisciplinary research initiative, bringing together several teacher-researchers from different academic branches with the ambition of creating operational artificial intelligence tools that incorporate, from the design phase, ethical principles for the development of artificial intelligence at the service of the general interest.
The Operational AI Ethics initiative brings together part of the work of four Télécom Paris research chairs related to Artificial Intelligence: NoRDF project, Data Science and Artificial Intelligence for Digitilazed Industry and Services, Explicability of Artificial Intelligence for Money Laundering and Digital Finance.
5 pillars of research
The Operational IA Ethics initiative conducts its work along 5 pillars of research:
- Bias and fairness who wants to answer the question, « At what point does bias become unlawful discrimination? »
- Explainability of algorithms which seeks, among other things, to establish a framework for defining the « right » level of explainability based on technical, legal and economic considerations.
- Liability which addresses both algorithmic liability (Who should pay for damage caused by an AI system?) and prospective liability (How to ensure that designers, operators and users of AI systems have an incentive to minimize damage to others in the future).
- Governance and Regulation, focusing on the AI regulatory ecosystem, which wishes to answer the question: « Which instruments in the regulatory toolbox ensure the best trade-offs between innovation and AI safety? ».
- AI and Public Interest which focuses on how to integrate public interest objectives and KPIs (Key Performance Indicators) into AI programmes managed by government and by private actors whose mission involves the promotion of the public interest.
>> Web page of the initiative: www.telecom-paris.fr/en/ai-ethics