12.5.2022 |
EN |
Official Journal of the European Union |
C 194/1 |
Opinion of the European Economic and Social Committee on ‘Developing artificial intelligence in European micro, small and medium-sized enterprises (MSMEs)’
(own-initiative opinion)
(2022/C 194/01)
Rapporteur: |
Marie-Françoise GONDARD-ARGENTI |
Plenary assembly decision |
25.3.2021 |
Legal basis |
Rule 32(2) of the Rules of Procedure |
|
Own-initiative opinion |
Section responsible |
Section for the Single Market, Production and Consumption |
Adopted in section |
13.12.2021 |
Adopted at plenary |
19.1.2022 |
Plenary session No |
566 |
Outcome of vote (for/against/abstentions) |
238/0/3 |
1. Conclusions and recommendations
1.1. |
The European Economic and Social Committee (EESC) points out that micro, small and medium-sized enterprises (MSMEs) will take up artificial intelligence (AI) to the extent that it is reliable, inclusive and sustainable, and if it forms part of an environmentally-friendly economic and social set-up — a factor in business development, employability and better quality of life and health for all workers and European citizens. Respect for fundamental and social rights and stronger transparency requirements will foster trust and ownership of AI by all citizens and MSMEs. |
1.2. |
The EESC underlines that the self-employed and micro, small and medium-sized enterprises across all sectors account for 99 % of EU businesses, two-thirds of private sector jobs and more than half of the added value created by EU enterprises. Exposed to a form of competition radically altered by the digitalisation of the economy, they nevertheless have a key role to play in the EU’s digital and green transition in order to meet the challenges ahead. They must therefore be given equal access to AI, as otherwise Europe will be deprived of its most valuable economic, social and human resources. |
1.3. |
The EESC pointed out in its opinion on the revised coordinated plan (1) that the widespread adoption of AI by MSMEs is the key to Europe’s industrial renewal. Despite their agility, MSMEs concerned about the digital transition face significant internal and external challenges: costs, lack of broadband infrastructure in certain territories, access to finance, human resources, information, training, etc. |
1.4. |
The EESC calls for MSMEs to be provided with simple tools and accessible funding to help them integrate this technology which, though expensive, is essential to maintain and even strengthen their competitiveness. Access to data of sufficient quality and quantity, as well as full-scale trials, is a priority. |
1.5. |
The EESC considers that effective support for MSMEs in adopting AI requires strong political will at all levels, close cooperation with all stakeholders in organised civil society and high-quality social dialogue in the Member States. |
1.6. |
The variety of MSMEs across different sectors requires tailored and targeted support for AI, as well as smart and inclusive legislation, which are key elements in ensuring legal certainty and trust. |
1.7. |
The EESC recommends the rapid dissemination among all stakeholders, and business managers in particular, of the ‘toolbox’ — an educational tool featuring in the EESC study on this subject (2) and describing the different stages in the use of AI by MSMEs. Such tools serve to dispel any anxieties about this technology and change perceptions in this area. |
2. General comments
2.1. |
The study Boosting the use of artificial intelligence in Europe’s micro, small and medium-sized enterprises, published by the EESC, analyses five sectors (agriculture, construction, healthcare, legal services and accounting) in five Member States (Italy, France, Ireland, Romania and Sweden). The recommendations for a ‘toolbox’ are achievable for policy-makers and MSMEs alike. |
2.2. |
The SME report 2021 published by the European Commission (EC) in July (3) clearly identifies the current situation in terms of digital uptake by MSMEs in the 27 EU Member States on the basis of two surveys, conducted during the first phase of the pandemic and the last quarter of 2020. |
2.3. |
Europe’s 25 million MSMEs, which are pillars of the EU economy and represent around 100 million jobs and almost 57 % of European GDP, play a key role in value creation across all sectors (from independent traditional crafts to the social economy, i.e. 2,8 million businesses, high-tech start-ups, etc.), and call for tailored and targeted support, as well as smart and inclusive legislation, ensuring legal certainty, trust and quality of life for workers. |
2.4. |
The EESC hopes that the measures taken by the Commission and the Member States will continue to be implemented at grassroots level. The above-mentioned toolbox should encourage public authorities to experiment with tailored initiatives in collaboration with civil society and relevant businesses. |
2.5. |
Ensuring the transformation of industries, AI often remains inaccessible to MSMEs: their agility comes up against difficulties both internal (human resources, costs, access to quality data, etc.) and external (access to broadband infrastructure, finance, information, training, etc.). |
2.6. |
The widespread adoption of AI by MSMEs is the key to Europe’s industrial renewal (4), as the most vulnerable, the self-employed, micro-enterprises, family structures, businesses far from urban centres as well as social economy enterprises deserve special attention to avoid discrimination that would have disastrous economic and social repercussions for the EU as a whole. |
2.7. |
Instilling a climate of trust is of the utmost importance for including MSMEs in the strategy of excellence and leadership strived for by the EU. This requires the mobilisation of all institutional actors, from the European Commission to national governments and regional and local structures, in order to provide tools and support targeted according to size and to ensure support and guarantees for investment. Bridging the digital divide and integrating networks to close the gap between MSMEs and larger entities in terms of AI are of key importance. |
2.8. |
Trust arises from a strong political will to work closely at all levels: organised civil society, social partners, associations, chambers of commerce, professional organisations, groupings, etc., which are on the ground, well known and recognised by both employers and by employees of MSMEs, are best placed to inform workers and employers about risks and challenges depending on the sector of activity. They have all the competences needed to inform policy decisions at all levels, so that these meet MSMEs’ needs in relation to AI. The EESC believes that it would be in the Member States’ interest to foster high-quality social and contractual dialogue in order to maximise the impact of policies aimed at MSMEs. |
2.9. |
The importance of effectively targeting European programmes at MSMEs (5), such as Digital Europe and Horizon Europe, is of obvious value. All tools must be mobilised in a form that is simple to use. |
2.10. |
The increasing sophistication of algorithms affects data quality and protection, and even security and transparency, raising many ethical concerns (discrimination, increasing social inequalities, doubts over the autonomy of human decisions, etc.). These challenges which require everyone, including MSMEs, to be highly vigilant, should not delay the uptake of AI by these enterprises. |
2.11. |
Many of the EESC’s recommendations are supported by the results of the above-mentioned study:
|
2.12. |
The EESC supports the recommendations of the above study, stressing that they require a particularly active and consistent political will:
|
2.13. |
The EESC would like to draw the attention of the European institutions to the following principles of action regarding AI:
|
3. Specific comments (13)
3.1. Agriculture
3.1.1. |
The agricultural sector, which features a wealth of family-run MSMEs (96 % of farms in 2016), faces a number of challenges: demographics, climate change, attractiveness and competitiveness. |
3.1.2. |
In view of the proven benefits of AI and robotics in this sector, it is advisable to provide continuous help to farmers and their employees in supporting these changes, to develop training in this area and appropriate education on the sharing and aggregation of data (14), while exploring new ways of sharing costs such as cooperative solutions. |
3.2. Construction
3.2.1. |
This labour-intensive sector could benefit more from digitalisation: 36 % of companies in the sector have adopted at least one AI technology, compared to 42 % across all sectors. |
3.2.2. |
Building information modelling (BIM), which is increasingly important for the sector and used by 29 % of enterprises in 2016, requires massive investment in training and in ensuring open and non-discriminatory access to electronic planning software via open and standardised interfaces (‘open BIM’), so as not to compromise market access for MSMEs. |
3.2.3. |
The characteristics of this sector and its potential could make it a showcase for a green and digitalised Europe if the prerequisites for securing market access are ensured in terms of technology and public procurement law. |
3.3. Health
3.3.1. |
In the field of health, more than elsewhere, innovation must meet ethical and safety requirements, from design to use. The development of an AI sector in the field of health requires access to sufficient and high-quality data. Territorial ‘sandbox’ experiments would be welcome to help MSMEs in this regard. |
3.3.2. |
Particularly active in medical equipment (95 % of the 32 000 companies in the sector) and services, MSMEs are at the forefront of innovation and productivity (47 % of companies have already adopted AI technology). |
3.3.3. |
However, MSMEs in the sector are confronted with the high cost of R & D and long innovation cycles, complex marketing authorisation procedures and the binding GDPR framework. |
3.3.4. |
The creation of a European space for reliable health data, as well as the recruitment of ‘hybrid’ employees with a mastery of both the job and digital techniques, would foster innovative solutions, including in MSMEs, as a source of savings for more sustainable and resilient care systems (15). |
3.4. Liberal professions (16)
3.4.1. |
Given the very sensitive nature of data held by liberal professionals, the stakes in terms of privacy, confidentiality, transparency and non-discrimination are particularly acute in this sector. |
3.4.2. |
The use of AI for automation of the most tedious tasks should allow liberal professionals to refocus on the core business of their profession to meet the needs of their clients/patients even more effectively. |
3.4.3. |
As highlighted in the EESC Opinion on Liberal Professions 4.0 (17), AI raises questions for the special relationship of trust between the liberal professional and his or her client, as well as the concepts of independence and professional responsibility. This entails reviewing the ethical rules by incorporating technical and ethical aspects related to AI and setting new training and skills requirements (data security and quality, personal data protection, cybersecurity, etc.). |
3.5. Legal advice
3.5.1. |
Despite sustained and independent growth through economic cycles (2,6 % per year between 2014 and 2018), progress in this sector has been slowed by the problems of transparency and bias that have affected AI since its inception in an area where confidence is based on direct human contact. |
3.6. Accounting/management control
3.6.1. |
This sector, with 640 000 companies operating in Europe in 2018, is dominated by MSMEs. Given that around half of accounting tasks could be automated, AI represents an opportunity in terms of added value and would allow professionals to refocus on providing advice to clients. |
3.6.2. |
The formation of hybrid profiles combining law or accounting with IT would be of great value in both sectors, in particular to prevent the risk of data piracy in the latter. |
Brussels, 19 January 2022.
The President of the European Economic and Social Committee
Christa SCHWENG
(1) OJ C 517, 22.12.2021, p. 56.
(2) EESC Call for tender No CES/FSA/02/2020; author: Space Tec Partners.
(3) Annual Report on European SMEs 2020/2021, European Commission, July 2021.
(4) Opinion on the revision of the Commission’s coordinated AI plan (OJ C 517, 22.12.2021, p. 56).
(5) EESC opinion (OJ C 429, 11.12.2020, p. 210).
(6) Announcement of 10 November 2021 by Commissioner Thierry Breton.
(7) See: OJ C 240, 16.7.2019, p. 51.
(8) As defined in recital 36 of the Commission proposal for a Regulation on data governance, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52020PC0767
(9) EESC opinion (OJ C 429, 11.12.2020, p. 1).
(10) Cf. points 4.7 and 4.8 of EESC opinion (OJ C 429, 11.12.2020, p. 1).
(11) See: OJ C 240, 16.7.2019, p. 51.
(12) See: OJ C 240, 16.7.2019, p. 51.
(13) These specific comments concern the different sectors analysed in the above-mentioned EESC study.
(14) Farmers, despite being major users of satellite data and automation, are apparently reluctant to share their data, according to the 2021 report on Artificial intelligence — state of the art and the outlook for France.
(15) Cf. in particular the EESC opinion published in OJ C 341, 24.8.2021, p. 76, the opinion published in OJ C 517, 22.12.2021, p. 56 on the AI coordinated plan, and also European Commission’s Strategic Foresight Report, 2021.
(16) Defined by the CJEU as follows: the liberal professions are ‘activities which, inter alia, are of a marked intellectual character, require a high-level qualification and are usually subject to clear and strict professional regulation’ (C 267/99, I-7467, 2001).