France AI Strategy Report

In March 2018, Emmanuel Macron, the President of the French Republic presented his vision and a 5-year national AI strategy. The French AI strategy is entitled AI for humanity (France, 2018a) and has been developed on the basis of the AI policy report (France, 2018b) prepared by French Member of Parliament (MP) and renowned mathematician Cédric Villani.

The main objectives of the French AI strategy as highlighted by the French President are to:

• Improve the AI education and training ecosystem to develop, retain and attract world-class AI talent;
• Establish an open data policy for the implementation of AI applications and pooling assets together;
• Develop an ethical framework for a transparent and fair use of AI applications.

To this purpose, the French Government will dedicate EUR 1.5 billion to the development of AI by the end of 2022, including EUR 700 million for research.

In 2021, the Prime Minister Jean Castex announced a renewed open and shared data strategy following a 2020 Report by MP Eric Bothorel which includes data and datasets for AI.

Also, the 4^th generation of the multi-year National Investment for the Future Programme is being framed in 2022. One sub-programme will be dedicated to AI and several other sub-programmes will include actions related to the National AI Strategy. This will allow updating the National AI Strategy funding and secure budget slots corresponding to the EU renewed Coordinated Plan.

France AI Policies on OECD.AI dashboard

In order to reduce the AI, data science and robotics skills gap on the labour market, the AI strategy will continue to provide financial incentives to higher education and research institutions to increase the provision of initial training at all levels, intermediate and expert, dual programmes and the retraining or upgrading of talent. Initiatives to increase diversity in computer science and AI are being deployed.

• The number of graduates at Masters’ level has doubled since 2016 and a new doubling target has been set;
• The range of training courses, in initial and part-time continuing education, summer or winter schools and executive programmes altogether has increased rapidly in 2019-2021, especially in the 4 Interdisciplinary Institutes of Artificial Intelligence (3IA) created in 2019: Toulouse-ANITI, Nice-Côte d’Azur 3IA, Grenoble-MIAI and Paris-PR[AI]RIE, but also in 3 other major education and research centres specialized on data science, AI and Robotics: Paris-Saclay-Institut data-IA, Institut Polytechnique de Paris-Hi! and Paris-Sorbonne-SCAI;
• Foster education of all players involved in the "algorithmic chain" (designers, professionals, citizens), including intermediate capacities for the working force and enable each citizen to improve its digital literacy to better understand the inner working of machines and the benefits of AI.

In terms of vocational training and lifelong learning initiatives, the French Government highlights:

• The formation of the Grande Ecole du Numérique (GEN): created to support training that helps to integrate people at risk of unemployment to the job market by developing their digital skills;
• The move towards quality labelling, certification and integration into rich professional training paths, including re-skilling and up-skilling, such as the efforts by the OPIIEC digital occupations observatory.

The deployment of AI technologies will have a major impact on the job market. According to France’s Employment Orientation Council roughly half of the occupations could be automated in the medium to long term. To tackle this issue, the French AI strategy devotes particular attention to a better understanding of future labour demand and skill needs to prepare successfully for professional transitions. The following policy recommendation is targeting increased labour market intelligence and forward-looking skills predictions:

• The creation of a public laboratory on the transformation of work to encourage reflection on the ways in which automation is changing occupations and to provide support for professional transitions.

The goals set by the National AI strategy extensively rely on applied research and innovation. The responsibility for coordinating the research side of the AI national strategy has been given to the French national research institute for the digital sciences (Inria), with a clear objective: to strengthen the entire French AI sector, accelerate technology spinoffs or transfers, and develop cooperation programmes with the industry. Among others, the research institute will coordinate the strategy’s implementation, provide scientific and technological expertise and develop bilateral cooperation initiatives, in particular with Germany.

The network of interdisciplinary AI research institutes within universities relies on some form of specialisation, directly linked to the regional academic and economic ecosystem. Incitation schemes have been developed, implying that 1 euro of private partners’ investment in collaborative R&D&I programmes results in 1 additional euro of state funding.

So far, the National AI strategy has helped to create 180 additional academic chairs and 300 additional PhDs in AI disciplines. Many other non-profit institutions help creating an innovation-friendly ecosystem and translating the strategy into reality:

• Several of the 16 Technology Research Institutes (IRT) or Energy Transition Institutes (ITE) help leading partnering projects in AI, among which IRT-System X on Trustworthy AI, IRT Saint-Exupery on Complex Systems embedding AI, IRT nanoelec on hardware components, ITE Efficacity in the Smart City domain, ITE Vedecom in the automotive industry. The association of IRT and ITE has developed a cross-institutes AI Programme, called EngageAI;
• Several of the 53 French sectoral poles de compétitivité (Innovation Clusters) also develop active AI ecosystem animation, with a view to disseminate AI in industry, such as Aix Safe Cluster (Safe city and security), Systematic-Paris-Region (Deep Tech), Paris-Cap-Digital, Medicen Paris Region (Health), Rouen Mov’eo (Mobility), and the Toulouse Aerospace Valley;
• The Teralab Institute provides technological resources and a whole ecosystem of experts companies, to remove scientific and technological obstacles faced by organisations wishing to exploit their data and accelerate experimentation and technology transfer.

The AI for Humanity strategy prioritises policy support for research and innovation to specific sectors that show sufficient maturity to embrace major AI transformations: health, transport, the environment, defence and security. Hence, this will require sector-specific policy, including sector-specific data platforms for data compilation and exchange, large-scale computing infrastructures and testing facilities. While many of these aspects touch upon infrastructure (cf. below), the French Government advocates creating test areas to facilitate the design and deployment of AI technologies:

• Implement test areas and innovation sandboxes to facilitate experimentation in real-life conditions while temporarily reducing the regulatory burdens to help testing innovations;
• The economic national operator bpifrance (public investment bank France) also operates yearly generic funding, prizes and labels schemes that target the digital transition at large but whose share of AI projects has been steadily increasing. It proves essential to consider AI innovation as an applied and integrative endeavour.

Following initiatives are envisaged to foster networks and collaborations in AI:

• Inria will coordinate the network of French AI expertise by means of the development of the 3IA Institutes and other research-oriented collaborative support mechanisms;
• Trilateral French-Japanese-German research projects on AI: The French National Research Agency (ANR) together with the German Research Foundation (DFG), and the Japan Science and Technology Agency (JST) is announcing the first trilateral call for research proposals on AI. This call is intended to support collaborative projects of trilateral research teams over 3 years, bringing together research partners from France, Germany, and Japan;
• Fostering of public-private laboratories, so called LabComs to encourage collaborative AI research and innovations;
• Inria and the DFKI, signed a memorandum of understanding aimed at building a strategic partnership between France and Germany on AI. The goal is to pool our strengths in order to overcome obstacles in the fields of health, cybersecurity, robotics and industry. More information about this collaboration is presented on page 11 of the 2020 activity report of Inria;
• Participating to the Global Partnership on AI (GPAI): a declaration has been signed between France and Canada to start a project on the creation of an international initiative to spur a responsible development and use of AI in full respect of human rights, inclusion, diversity, innovation and economic growth. The GPAI collaborates with international partners and organisations to bring together experts from industry, civil society, governments and the academic world. This initiative is stirred by a secretariat, hosted by the OECD in Paris, and it accounts for two Centres of Expertise in Montreal and in Paris.

To foster the international attractiveness of AI in France, the French strategy expresses the need for policies to boost France’s appeal to expatriates and foreign talent by improving working conditions and salaries of researchers.

Ethical matters to ensure a fair and transparent use of AI technologies and algorithms are central to the French AI strategy. In this regard, Cédric Villani’s recommendation in the AI policy report to create a “digital technology and AI ethics committee in charge of leading public discussion in a transparent way, and organised and governed by law” has effectively led to the creation early 2020 of a Pilot National Digital Ethics Committee (CNPEN). In the pilot phase, it has been tasked about 3 areas of AI ethics but will extend progressively its outreach.

The French strategy highlights the following data policy initiatives:

• The AI-specialized high-performance computer infrastructure Jean ZAY, inaugurated in early 2020 and currently running at 28 petaflops capacity (X2 since inauguration);
• The CASD secure Data Hub, a public interest group bringing together the State represented by the French National Institute of Statistics (INSEE), Groupe des Écoles Nationales d'Économie et Statistique (GENES), the French National Centre for Scientific Research (CNRS), École Polytechnique and HEC Paris to help exchange securely sensitive in protected data for R&D projects;
• Data sharing in private sector: the French Government has funded many AI Challenges (3^rd season 2020) to help sponsors tackle data valorisation and problem resolution through AI in cooperation with start-ups and other innovators, and also funded several data hub projects to foster the setting up and development of sectoral data spaces (namely in agriculture, alimentation, logistics, health, sport), to be fully integrated in European data spaces;
• Encourage the creation of data commons fit for the AI age, including the provision of open data sets, and more real-time open data. This perspective is reflected in the National plan for open science;
• Increased data portability: the right to data portability should be supported, allowing migration of data from one service ecosystem to another without losing data history.

In terms of digital and telecom infrastructure to encourage the development of machine learning and AI algorithms, the French strategy strongly supports and participates in the following policy initiative:

• The GAIA-X project, initiated by Germany and France, is to create a secure, federated data system that meets the highest standards of digital sovereignty while promoting innovation.

Climate and environment

The AI policy report, prepared by French MP Cédric Villani, includes a chapter entitled Using artificial intelligence to help create a more ecological economy with recommendations to:

• Fostering greener AI: within this area, public authorities are called to support a greener value chain and the ecological transition of the European cloud industry;
• Promoting the dissemination of ecological data: open ecological data (i.e. data on weather, agriculture, transport, energy, biodiversity, climate, waste, land registry and energy performance assessments) are the key point around which green AI technologies can develop and promote ecological transition.

Another AI policy report, published in 2019, on the prospective of AI in France commissioned by the Ministry of the Territorial Cohesion and the Ministry of Economy and Finance, presents a similar signal. It highlights the need to strengthen the national ambition for the use of AI to the benefit of the energy sector and the environment. Among others, it calls for the development of intelligent energy networks through the use of Smart grids (based on smart meters and appliances), and new generations of supervisory control and data acquisition (SCADA) with high levels of digitalisation and remote control interventions. A cutting-edge infrastructure can enhance the deployment of AI applications for the optimisation of energy consumption, a higher quality service on the network, and better energy storage capacities.

With respect to the environment, the use of AI could be beneficial in the following areas:

• Reduction of the use of natural resources;
• Anticipating environmental risks;
• Increasing the efficiency and integration of renewable energy sources.

Lastly, researchers in optimisation and AI at Inria Lille - North Europe have successfully completed Perf-AI, a research project executed in the framework of the European H2020 programme. Carried out in collaboration with a start-up from the aeronautics sector, and based on the analysis of flight data, the project led to the development of digital aircraft models enabling the optimisation of flight plans. Perf-AI thus offers a solution to reduce the energy consumption of commercial airlines.

COVID-19 pandemic

In terms of AI-related policies to response specifically to the COVID-19 pandemic, following collaborative initiatives are ongoing or in preparation:

• Within the framework of the GPAI a working group on AI and pandemic response (AIPR) has been formed to promote cross-sectoral and cross-border collaboration in this area. In November 2020, the working group released a report outlining its mandate and providing recommendations to foster and support the responsible development and use of AI-enabled solutions to address COVID-19 and future pandemics;
• The French Ministries of Social Affairs and Health Education and of Research and Innovation are supporting 20 research projects to fight against the COVID-19 of which one employs AI to reconstitute the COVID-19 replication process, to model it, and to finally test suitable inhibitors.

Furthermore, in the fight of the COVID-19 pandemic, the French National Institute of Health and Medical Research (INSERM) and researchers at the University of Paris created a national phone line operated by intelligent digital assistants. The name of this service is AlloCOVID and it operates with AI–based virtual phone assistants that can respond to more than a thousand people at the same time (24/7). In addition, French authorities have integrated new AI tools in the security cameras of the Paris metro system, which could be expanded to other public transport means. The software was developed by French start-up DatakaLab to create anonymous data that will help authorities anticipate future outbreaks of COVID-19.

Inria, the French national research institute for the digital sciences, has committed to play a prominent role as coordinator of the national AI strategy. It is responsible for its implementation, in particular on its research and innovation side.

Launched in 2018 the first phase of the French national strategy in AI, with a budget of EUR 800 million for 3 years, put a strong emphasis (1/3 of spending) on boosting research, with the creation of interdisciplinary institutes 3IA, the extra financing of 180 PhDs and the opening of a petascale supercomputing facility.

The second phase of this strategy (2021-2022) sets main priorities on the upscaling of education and training, development of embedded AI and trustworthy AI in critical systems in order to strengthen the national industrial base while accelerating the digital and ecological transition of companies thanks to AI.

France (2018a). AI for Humanity: French Strategy for Artificial Intelligence. President of the French Republic. https://www.aiforhumanity.fr/en/

France (2018b). For a Meaningful Artificial Intelligence: towards a French and European Strategy. Cédric Villani, Member of the French Parliament. https://www.aiforhumanity.fr/pdfs/MissionVillani_Report_ENG-VF.pdf

France on OECD.AI observatory 

Read full report