The EU’s AI Act is the first comprehensive legal framework regulating artificial intelligence globally. But regulation alone is not going to make us competitive in this field. Instead, Europe must invest in secure connectivity and resilient backbone infrastructure as well as a resilient supply chain. Europe has seen considerable advancements in generative AI. The pieces of the puzzle for a European value chain are there – European companies that provide cloud services, who run data centres; top-level researchers and cutting-edge research facilities for chips; start-up capital for talented entrepreneurs; academic-industry collaboration; user demand, especially from industry for innovative general-purpose models and applications.
Take-up of Generative AI is already there in the pharmaceutical industry, in manufacturing, media, architecture, interior design, engineering, automotive, aerospace, defense, medical, electronics and energy industries. All these sectors are augmenting core processes with AI models. The technology does impact marketing, design, corporate communications, and training and software engineering by augmenting the supporting processes that span many organizations.
Of course, there are risks to AI and General Purpose AI. AI applications rely on models, which need to be trained. Successful training relies on data – enormous amounts of data. The way the training is done influences the quality of the outcome, of the trained model or AI application. There can be flaws in the models and bugs in the applications. And there is obviously also outright criminal use of the technology. Thus, the EU also must have the expertise to refute malicious attacks and cyberthreats, it must be able to rely on EU-based infrastructure, to ensure that, to put it simply, “the lights stay on”.
It shows the importance of having the right regulation in place, based on European values, such as transparency, sustainability, data protection, respect for the rule of law. The EU’s AI Office, as introduced through the AI Act, and national authorities play, from the perspective of the EESC an important role in this context. They must have the necessary resources to monitor, evaluate and enforce the provisions of the law and ensuring compliance. This is not an easy ask – many of the major GPAI developments are currently being spear-headed by actors outside the EU’s jurisdiction. Being aware of such relevant developments and having the capacities to enforce their compliance with its regulatory provisions requires specific resources. As generative AI is constantly evolving, frequent updates with regular feedback from the scientific community and stakeholders will be needed. While the voluntary codes of practice with respect to the AI Act could make it easier for companies to comply with the regulation, this will only happen if they are introduced quickly and in a user-friendly way, with easy and speedy certification processes. Otherwise, they will only create an additional burden.
As representatives of social partners and civil society, we strongly advise for the European Commission to organise dialogues with stakeholders, including social partners, about the codes of practice in workplaces and workers’ rights in the context of GPAI to dispel fears, ensure broad understanding of the new technology among citizens, support for its uptake and awareness of the risks and how to avoid them. We also call for support for European companies active in the GPAI field, ranging from an enabling legal framework to active procurement of their products by public authorities in the EU. This will go a long way in promoting European solutions also beyond the public sector, demonstrating their trustworthiness to further users and clients. For example, cloud hosts have unparalleled power to monitor, detect and stifle competitors. Currently, the main providers are US-based companies. But European cloud service providers can be part of a solution that meets the demands for compliance with European values and security needs and should be promoted, e.g. through using their services in EU public procurement projects.
Furthermore, in the international context, the EU needs to use its voice in international standard setting organisations to highlight and support European technologies. As many countries are searching for ways to control AI developments, rapidly establishing and applying standards for evaluating GPAIS is of global interest. The EU is in a pole position, since it can already point to the AI Act and other related regulation. It must now ensure that the international developments in this area are aligned with our provisions. Ultimately, the EU has another powerful tool at its disposal – by ensuring that competition authorities in the EU leverage their capacities, they can make certain that so-called hyper-scalers do not abuse their B2B or B2G market position.
The to-do-list for the new European Commission and the co-legislators is clear:
- ensure resources for monitoring and enforcing the EU legislation related to AI are put in place
- provide and attract public and private funding for scaling-up efforts by entrepreneurs,
- invest in the critical digital infrastructure, including cybersecurity
- combat market concentration from non-European digital companies
- overcome the continuous fragmentation of the internal market which hinders scaling-up and impedes the competitiveness of Europe’s AI actors.
Currently, the EU’s GPAI developers and users are overly reliant on US-based cloud and training data providers. Given the changing geopolitical situation, this is not a sustainable situation. While legal and technical cooperation remains important, the EU must develop its own secure and high-performing digital infrastructure and AI ecosystem. And as the EU’s vision on AI includes, unlike the stance favoured by the US, bans on the most invasive forms of the technology and strict rules requiring hyperscale companies to be more open about how they design AI-based products, it has to get its own act together.
Sandra Parthie, president of the Single Market, Production and Consumption section in the European Economic and Social Committee