Access to European Union law
<a href="https://eur-lex.europa.eu/content/help/eurlex-content/experimental-features.html" target="_blank">More about the experimental features corner</a>
Choose the experimental features you want to try
EUR-Lex
Access to European Union law

This document is an excerpt from the EUR-Lex website

You are here
Use quotation marks to search for an "exact phrase". Append an asterisk (*) to a search term to find variations of it (transp*, 32019R*). Use a question mark (?) instead of a single character in your search term to find variations of it (ca?e finds case, cane, care).
Search tips
Need more search options? Use the Advanced search

Document 52021AE2456

Opinion of the European Economic and Social Committee on Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions — Fostering a European approach to Artificial Intelligence (COM(2021) 205 final)

EESC 2021/02456

IO C 517, 22.12.2021, p. 56–60 (BG, ES, CS, DA, DE, ET, EL, EN, FR, HR, IT, LV, LT, HU, MT, NL, PL, PT, RO, SK, SL, FI, SV)

22.12.2021   

EN

Official Journal of the European Union

C 517/56

Opinion of the European Economic and Social Committee on Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions — Fostering a European approach to Artificial Intelligence

(COM(2021) 205 final)

(2021/C 517/08)

Rapporteur:

Marie-Françoise GONDARD-ARGENTI

Referral

European Commission, 1.7.2021

Legal basis

Article 304 of the Treaty on the Functioning of the European Union

Adopted in section

2.9.2021

Adopted at plenary

22.9.2021

Plenary session No

563

Outcome of vote

(for/against/abstentions)

235/0/7

1.   Conclusions and recommendations

1.1.

 The European Economic and Social Committee (EESC) welcomes this revised coordinated plan on artificial intelligence (AI), resulting from collaboration between the European Commission and the Member States, and calls for its expedited implementation by all stakeholders playing their part.

1.2.

 The EESC supports the Commission’s holistic approach to its strategy for inclusive and sustainable AI, based on the creation of an ecosystem of excellence and trust. The proposals for plans and regulations, each the subject of a separate opinion, are the two pillars for positioning Europe as a global player in AI.

1.3.

 In order to yield the expected economic, social, environmental and public health benefits for all of Europe’s citizens without distinction, the plan must inject new momentum into education by incorporating ethical and environmental issues, lifelong learning and skills development, as well as innovation challenges and support for workers and businesses, including smaller ones.

1.4.

 In the world of work, trust is the key to success, and AI is no exception to this rule. The EESC believes that this new plan should give social dialogue and the social partners a key role throughout the process. While the question of skills is essential, consideration must be given to all issues pertaining to automation in the world of work.

1.5.

 The EESC recommends that the Commission, within the policy framework it intends to establish, strongly encourage Member States to strengthen social and civic dialogue on AI issues and tools.

1.6.

 The EESC points out that trust hinges on the right of every worker, consumer or citizen to the explicability of algorithmic processes in so far as these have an impact on their lives and their environment. The new coordinated plan must be part of this requirement of transparency and explicability, inseparable from the right of every citizen to challenge decisions taken solely by an algorithm.

1.7.

 The EESC believes that the EU can establish itself as a credible global leader in AI provided there is unwavering coordination between EU institutions, Member States and organised civil society — social partners, consumers and citizens — with a clear definition of individual responsibilities. By operating between European institutions, between them and Member States and between Member States, this coordination will maximise the overall impact while ensuring the rapid and consistent implementation of investment policies and programmes.

1.8.

 While AI is intrinsically global in nature, the scale of its deployment is a decisive factor. By elevating Member States to key players on the global AI stage, the coordinated plan is a cornerstone of the European AI strategy.

1.9.

 The EESC welcomes the fact that the revised coordinated plan proposes specific measures to boost the implementation of the national strategies and increase synergies. However, the plan only takes on its full significance in tandem with the European data strategy — the creation of a European space for data and data protection — and the strategy for cybersecurity.

1.10.

 To overcome the lack of binding coordination, the EESC calls for a scheme providing for ongoing monitoring of progress using performance indicators devised by all stakeholders.

1.11.

 The EESC encourages the Member States to use the Recovery and Resilience Facility — as well as other financial instruments such as Horizon Europe, the Digital Europe programme and InvestEU — to support the development and uptake of AI, especially among the smallest businesses. A recent EIB report sheds light on the road ahead (1).

2.   General comments

2.1.

 The Commission intends to make the EU the global driving force for human-centric, inclusive, safe, sustainable and trustworthy AI. Drawing lessons from the pandemic, it wants to speed up investment, strengthen the current measures and align them rigorously to foster synergies and agility (2). The EESC endorses this course of action aimed at bolstering the EU’s internal and international excellence and would like to see it rolled out with real urgency and resolve.

2.2.

 The EESC notes the progress made since the first coordinated plan in 2018. The updated plan clearly sets out the priorities: investing in infrastructure and data-sharing, fostering excellence by harnessing innovation and skills, and ensuring trustworthiness and transparency across the value chain. The investment of EUR 7 billion in 2019 (+ 30 %) is encouraging.

2.3.

 The coordinated plan is designed to ensure that algorithms comply with fundamental and social rights and the right to non-discrimination. This opinion goes hand in hand with the opinion on the related proposal for a regulation dealing with the ethical aspects and establishing harmonised rules on AI (3).

2.4.

 The EESC believes that the ethical framework and recognised and shared norms, together with a coordinated plan, are the key, defining catalysts providing the confidence for Europeans’ commitment to AI while ensuring a competitive advantage and an asset at global level.

2.5.

 This trust on the part of citizens is inextricably linked to the right to the explicability of algorithms and the possibility of challenging any decision affecting their lives. Developers of AI and its applications, as well as users, must be vigilant, with regular audits and testing. People must remain in control of decision-making.

2.6.

 The EESC is concerned about the risk of fragmentation linked to the different national strategies and calls on the Member States to step up their cooperation, particularly on cybersecurity, a sovereign prerogative that nevertheless forms the backbone of the plan, in the interest of the EU.

2.7.

 The EESC calls for sustainable AI when it comes to AI’s economic, social and environmental objectives and at the same time its economic, social and environmental responsibilities.

3.   Specific comments

3.1.

 Fully justified, the Commission’s objectives require it to work together with the Member States, the private sector and organised civil society to:

—

accelerate investment in AI with a view to a rapid and resilient economic and social recovery, based on advanced digital solutions;

—

deploy the planned structures and programmes in a timely manner to reap the legitimate rewards of first-movers in the production and adoption of technologies;

—

optimise AI policy alignment to remove fragmentation and act together to address a global challenge.

3.2.

 The EESC endorses the Commission’s intention to invest at least EUR 1 billion per year between 2021 and 2027 from the Digital Europe and Horizon Europe programmes: the projected leverage effect will propel public and private investment in AI to a level of EUR 20 billion per year over that period. The Recovery and Resilience Facility (RRF) will bolster this leverage effect.

3.3.   Creating favourable conditions — data and infrastructure

3.3.1.

 The EESC calls on the Commission and the Member States to step up cooperation and coordination within AI governance when it comes to implementation of the national strategies, and the sharing of knowledge, expertise and best practices to accelerate the development and uptake of AI, and thus rise to the global challenges of AI.

3.3.2.

 Availability, accessibility and free flow of data are the prerequisites for the development of AI, tying in with the strategy on data, data protection and cybersecurity. The EESC points out that the smooth exchange of personal data depends on close coordination and full compliance with the GDPR. As with data security and transparency, norms can provide a competitive advantage in global competition (4). The EESC supports the development of the European data space and of sector-specific spaces.

3.3.3.

 Apart from some AI systems that do not rely on data, the performance of AI depends directly on the volume, relevance and quality of the data available. The EESC therefore considers it essential to maximise this potential to facilitate flexible, interoperable and secure data exchanges. The EESC supports singling out key sectors such as industrial production, the Green Deal, mobility, health, finance and energy, so as to carefully monitor developments and assess the progress made.

3.3.4.

 The EESC supports:

—

the objective of significantly increasing the computing power available in Europe for the purposes of using and developing data in real time and the objective of producing the requisite processors in the EU as part of a sustainable development agenda;

—

the concept of data of general interest in the sense of public authorities opening up data that is helpful to them or to promote the sharing of data deemed essential (5), combining it with data from players operating in Europe but not involved in its industrial policy.

3.3.5.

 The EESC expects to see greater involvement of the Member States in research and innovation aimed at developing the structures, infrastructure and tools needed for the common good and for the purposes of contributing to the EU’s AI sovereignty.

3.4.   Fostering excellence from the laboratory to the market — innovation and partnerships

3.4.1.

 The EESC supports the holistic approach aimed at making the EU a place where excellence thrives from basic research and innovation to placing on the market. The EESC considers it essential that the EU and the Member States accelerate their cooperation in order to create the conditions conducive to innovation, AI business opportunities, and the development of talent and skills in line with Europe’s human-centred and trust-based vision of AI. The EESC is pleased to note that the cross-cutting collaborative structures support these objectives.

3.4.2.

 The EESC welcomes the Commission’s objectives of promoting experimentation, strengthening regional digital innovation hubs in cooperation with national and regional initiatives, and sharing experiences. These innovation hubs and on-demand platforms help VSEs and SMEs to take ownership of AI, in addition to targeted financial support. The EESC considers it important to support start-ups, scale-ups and traditional SMEs in AI development.

3.4.3.

 The EESC stresses, however, that this policy of centres of excellence must not undermine the normal and long-term funding of ‘standard’ research laboratories, for fear of delaying the emergence of disruptive technologies.

3.4.4.

 The EESC would like to promote programmes for organisations and MSMEs aimed at encouraging AI take-up, by developing targeted support networks. The initiatives put in place to support MSMEs’ access to AI (the testing and experimentation facilities (TEFs), Digital Innovation Hubs (DIHs) and AI-on demand platform) should be strengthened and assessed as to their practical impact on such businesses, with a view to improving them.

3.4.5.

 A recent study, commissioned by the EESC, showed the extent to which scaling-up is a prerequisite when it comes to access to the required skills, to data pools for algorithms and to the necessary funding, and with regard to the targeting of a market that is sufficiently developed to ensure the rapid amortisation of costly infrastructure. Policy-makers must therefore prioritise MSMEs and put forward tailored policies. In view of the economic damage caused by the pandemic, it is now imperative to support start-ups, scale-ups and traditional SMEs in AI development.

3.4.6.

 The EESC supports the development of ‘sandboxes’ to test new ideas in conditions close to reality and foster agility. The EESC calls for better sharing of results and their mutual recognition without borders.

3.4.7.

 Research excellence reflects the talents and skills that are found in laboratories, but also users’ trust in AI. The EESC therefore wants higher education programmes to generate world-class talent in innovation and commercial development. It calls on the EU to provide the conditions for retaining talent and attracting skills from abroad. The EESC calls for a stable and trusted global business environment to enable companies and workers to take greater ownership of AI for the benefit of research and innovation.

3.4.8.

 Civil society organisations deserve to be supported in implementing AI solutions and to be among the beneficiaries of AI funds and resources. This is particularly the case with social services, in full compliance with the right to privacy and the rules on the processing and storage of data, including biometric data.

3.5.   Making AI work for people — skills and trust

3.5.1.

 The EESC supports the Commission’s strategy based on the creation of an ecosystem of excellence and an ecosystem of trust. This approach is the subject of the White Paper adopted in 2019 (6). As education and training hold the keys to nurturing excellence and increasing trust in AI, it falls to education and training systems to meet a wide range of needs, from providing a basic knowledge and general understanding to helping people acquire specialised expertise and high-level skills (7).

3.5.2.

 The EESC welcomes the Commission’s adoption of the new Digital Education Action Plan for the period 2020-2027. It will help people to gain a better understanding of what AI is and how it works starting from primary school up to university, and to update their skills.

3.5.3.

 The EESC would like to see the Member States develop AI education that also incorporates the related ethical and environmental issues, and encourages the exchange of best practices in order to promote diversity in access to training, development and use of AI. If AI is to flourish and be used for the benefit of all, a multidisciplinary approach will be needed, with linkages forged.

3.5.4.

 The EESC considers the social partners to be key players in anticipating change in skills and jobs. Social dialogue is essential in helping to support workers affected by the automation of their tasks.

3.5.5.

 The EESC stresses the importance of cooperation between governments, educational institutions, the social partners and relevant civil society organisations in designing and rolling out new education and continuous training programmes, aimed primarily at people who are unemployed. Tailored training for MSMEs and SMEs is needed, for both entrepreneurs and employees.

3.5.6.

 The EESC regrets the fact that the plan fails to provide for a possible joint discussion on the world and future of work in the face of automation. Building workers’ trust in AI applications requires more rigorous social dialogue, which is needed when introducing AI systems that have an impact on workers, particularly in the areas of management and human resources. Given the potential for job losses, increased inequalities and restructuring of the productive sectors, seamless coordination is needed between the Member States, in close collaboration with businesses, regions and cities, the social partners and intermediary bodies, with a view to:

—

developing forward-looking, long-term automation capacities on an institutional footing;

—

developing innovative and proactive schemes for the transition of occupations and skills;

—

exploring complementarity, not only for ‘developing AI’ in occupations but also as regards tracking change in terms of culture and practice and establishing ‘good complementarity’ standards for individual empowerment.

3.5.7.

 As trust is the cornerstone of a digital society, the EESC particularly welcomes the Commission’s approach in all of the areas that can uphold the general interest and in particular the interests of citizens and consumers, as well as of workers and businesses, including in the social economy, so as to raise the overall level of trust in AI: the protection of fundamental rights, cybersecurity, data protection, intellectual property and the sustainable and efficient use of resources, without overlooking regulatory issues related to innovation. It seems important to pinpoint the possible dangers of the misuse of AI and to propose appropriate remedies. The EESC supports the measures envisaged by the Commission to strengthen security in all areas and adapt the regulatory framework to the challenges posed by AI. It points here to organised civil society’s pioneering role, for example in assessing the risks inherent in facial recognition.

3.5.8.

 The EESC fully supports the goal of promoting Europe’s vision for sustainable and trustworthy AI worldwide by encouraging the setting of global standards on AI and developing partnerships and joint initiatives with third countries.

3.6.   Building strategic leadership in high-impact sectors

3.6.1.

 The EESC welcomes the Commission’s realistic approach of focusing its action on sectors where AI leadership is essential: climate and the environment, health, robotics and automation, public services, public security including counter-terrorism and migration policy, smart mobility, agriculture and related sectors.

3.6.2.

 The EESC considers the digital transition to be a key instrument for decarbonisation and environmental management, while pointing to the need to limit the potential adverse impact of AI-based solutions on the climate, the environment and energy consumption. The use of AI in the management of energy and transport systems, as well as in industrial processes and agriculture, promotes energy efficiency, the circular economy and the sustainable management of natural resources; this contributes to productivity and economic efficiency. AI also provides the means to better understand and get to grips with climate and environmental phenomena.

3.6.3.

 The EESC calls for the opportunities offered by AI to be harnessed for the purposes of research into the origins of diseases, the development of new drugs, medical equipment and advanced care, and in the areas of investigation, diagnosis and treatment, in the fight against cancer for example. The European health data space, designed to ensure full respect of privacy and personal data, would be a major step in ensuring that AI’s potential is fully harnessed. Daily assistance to patients, older people and people with disabilities provided by robots illustrates the benefits of robotic excellence less concerned with technological prowess and more focused on mastering the interface between humans and robots.

Brussels, 22 September 2021.

The President of the European Economic and Social Committee

Christa SCHWENG

(1)  Artificial intelligence, blockchain and the future of Europe.

(2)  OJ C 240, 16.7.2019, p. 51.

(3)  EESC opinion on the Regulation on artificial intelligence (see page 61 of this Official Journal).

(4)  Norms over force. The enigma of European power, Zaki Laïdi, 2005.

(5)  Regulation on data governance.

(6)  OJ C 364, 28.10.2020, p. 87.

(7)  OJ C 14, 15.1.2020, p. 46.

Top