—

adoption of AI by MSMEs depends primarily on their managers’ and employees’ capacity to use it, including in relation to the risks, in advance of the deployment of these technologies;

—

only appropriate funding will compensate for the disadvantages in terms of size and maturity that de facto hamper the growth of AI in MSMEs. In this regard, the EESC welcomes the EUR 1,98 billion announced by the European Commission under the Digital Europe Programme (6). Innovation hubs offer the best access to the latest digital capacities and training of stakeholders;

—

MSMEs need high-quality external databases. The EESC has already stressed the importance of ensuring universal access to public sector big data, establishing robust programming interfaces (APIs) and even European platforms for secure data sharing (7), such as Gaia-X. It has also proposed a mutualised approach to data management and exchange and stressed the importance of developing altruistic data (8);

—

for the provision of AI technology, access to data, standardisation and access to finance, MSMEs need clear and uniform rules across the European market, which are conditions for healthy competition, in turn generating growth and jobs. In addition, they must participate as early as possible in the development of standards and regulation at European level. Clear and transparent solutions to ethical and accountability issues will make it possible to gain the trust of citizens and consumers, encouraging MSMEs to adopt AI.

—

to use education and vocational training to promote a general mastery of AI by citizens, so that all parts of civil society, including MSMEs, are able to freely adopt it, and act in a competent and responsible manner in this connection. It is essential to complement academic training with practical arrangements and to provide suitable and affordable continuing training for self-employed workers, managers of MSMEs and their employees (9). With regard to managers, their access to external expertise should be facilitated through B2B partnerships, an IA accessible to all, and the availability of AI as a Service;

—

to ensure that MSMEs have easy access to both public and private funding and to generate synergies for them in EU funding;

—

to guarantee the necessary infrastructure and connections in all areas, including rural areas, in order to prevent a digital divide and provide access to relevant and interoperable data, of which agriculture in particular is a major user;

—

to ensure good coordination between all actors and levels;

—

to promote general awareness of cybersecurity issues (10), as well as the economic and societal damage caused by biased data and other potential emerging risks. Human diversity in the design of AI tools is an essential means of improving data quality;

—

to widely disseminate best practices and success stories on the ground, facilitating the sharing of feedback to kick-start the widespread adoption of AI by MSMEs.

—

to start small, in order to inspire adequate and proportionate policies, even if it means adapting them for larger companies at a later stage;

—

to increase legal certainty and promote a good understanding of the rules in order to facilitate uptake by businesses and support investments to develop innovation and AI;

—

to facilitate understanding of AI by all through better coordination and increased synergies between policy instruments and initiatives, using a specific MSME platform;

—

to promote full-scale trials upstream through operational engagement protocols, among other things. The EESC highlights the importance of testing sites and regulatory ‘sandboxes’ that enable experimentation and piloting of new ideas (11) as well as the need to evaluate these tools in order to better adapt them (12). The digital innovation hubs, Enterprise Europe Network (EEN) and on-demand platforms need to work in close coordination to support SMEs in their implementation efforts;

—

to promote a pan-European approach by completing the European Single Market and adjusting policies to the needs of the Member States, from the development of AI to its adoption;

—

to allow open data access while respecting data protection and ownership and increasing data flows to feed AI-based systems;

—

to guarantee a transparent European Single Market to reduce risks and increase transferability of AI solutions to MSMEs. The interoperability of essential digital infrastructures is a key condition.