Report on Artificial Intelligence in a Digital Age
The European Parliament adopted by 495 votes to 34, with 102 abstentions, a resolution on artificial intelligence (AI) in the digital age.
The EU as a global reference
Parliament noted that the world is on the threshold of the fourth industrial revolution and that the ongoing digital transformation, in which AI plays a key role, has triggered a global competition for technological leadership. The EU has so far lagged behind, especially in comparison to China and the US, so that future technological standards risk being developed without sufficient EU input, often by undemocratic actors, which poses a risk to political stability and economic competitiveness. The EU should act as a global standard-setter on AI.
Members pointed out that by 2030, AI is expected to contribute more than EUR 11 trillion to the global economy. They stressed that the EU has the opportunity to shape the international debate on AI and to develop common global rules and standards, promoting a human-centred, trustworthy and sustainable approach to AI, in full respect of fundamental rights.
At present, the EU is still far from achieving its aspiration to become globally competitive in AI. Parliament therefore believes that the possibility of consolidating a distinctive European approach to AI on the international stage requires the Union to rapidly agree on a common strategy and regulatory framework for AI.
Reiterating the EU's call for a global agreement on common standards for the responsible use of AI, Members believe that the development of international technological norms and standards requires closer coordination and cooperation with like-minded democratic partners.
A clear regulatory framework
Parliament called for a regulatory environment for AI that ensures effective governance and protection of fundamental rights, while facilitating competitive access to digital markets for players of all sizes to promote innovation and economic growth for the benefit of all. It stressed that a competitive, accessible and fair data economy, based on common standards, is a prerequisite for the development of AI.
EU Roadmap up to 2030
With a view to making the EU a global leader in AI, Parliament presents its EU Roadmap for AI with clear policy recommendations for the next years. These include:
(1) Improving the regulatory environment
Members called on the Commission to only propose legislative acts in the form of regulations for new digital laws in areas such as AI, as the digital single market needs to undergo a process of genuine harmonisation. They called for consistent EU-wide coordination, implementation and enforcement of AI-related legislation.
Members called for consistent EU-wide coordination, implementation and enforcement of AI legislation. Digital legislation should always be flexible, principle-based, technology-neutral, future-proof and proportionate, while taking a risk-based and fundamental rights-based approach where appropriate and avoiding unnecessary additional administrative burden for SMEs, start-ups, academia and research.
The resolution highlighted that an underlying objective of the EU’s digital strategy, as well as that of the AI strategy, is creating a ‘European Way’ in a digitalised world. This approach should be human-centric, trustworthy, guided by ethical principles and based on the concept of the social market economy. The individual and the protection of their fundamental rights should always remain at the centre of all political and legislative considerations.
Members are convinced that it is not always AI as a technology that should be regulated, but that the level of regulatory intervention should be proportionate to the type of individual and/or societal risk incurred by the use of an AI system. They underlined, in this regard, the importance of distinguishing between ‘high-risk’ (which needs strict additional legislative safeguards) and ‘low-risk’ (which may, in a number of cases, require transparency requirements for end users and consumers) AI use cases.
(2) Completing the digital single market
Parliament called on the Commission to continue its work to remove the main unjustified barriers to the full completion of the digital single market. Members are convinced that the current national and European competition and antitrust frameworks need to be reformed to better target abuses of market power and algorithmic collusion in the digital economy, as well as data accumulation issues, and to better address the risks of new emerging monopolies without compromising innovation.
(3) Digital green infrastructure
Parliament called on the Commission to follow through on its ambition of incentivising 75% of European businesses to adopt cloud computing, bigdata and AI services by 2030 in order to remain globally competitive and accelerate its climate neutrality goals. It stressed that a functional and rapid infrastructure for AI must be based on fair and safe high-speed digital connectivity, which calls for the deployment of 5G in all urban areas by 2030. The EU is invited to take the lead in making green digital infrastructure climate neutral and energy efficient by 2030.
(4) Ecosystem of excellence
Parliament called on the Commission to create an AI skills framework for individuals and called for investment in research to better understand the structural trends related to AI in the labour market. It called on the EU to increase investment in research into AI and other key technologies, such as robotics, quantum computing, microelectronics, the internet of things, nanotechnology and 3D printing. It called for the extension of the digital Europe programme, considering that its funding of EUR 7.6 billion should be increased.
(5) Ecosystem of trust
Parliament called for awareness-raising campaigns to help citizens better understand the opportunities, risks and societal, legal and ethical implications of AI in order to make AI trustworthy and promote its democratisation. It called on Member States to put citizens at the centre of the design of online public services. It also called for an evidence-based approach to AI in health that prioritises personalised, patient-centred, cost-efficient and quality healthcare, while upholding human oversight and decision-making.
(6) Mass surveillance, military concerns
Members noted with concern that such AI technologies pose crucial ethical and legal questions. Certain AI technologies enable the automation of information processing to an unprecedented scale, which paves the way for mass surveillance and other unlawful interference and poses a threat to fundamental rights, in particular the rights to privacy and data protection.
Members called on the Commission and Member States to prioritise funding to AI research that focuses on sustainable and socially responsible AI, contributing to finding solutions that safeguard and promote fundamental rights, and avoid funding programmes that pose an unacceptable risk to these rights, including funding systems of mass surveillance, social scoring and other systems that have the potential to lead to negative social impacts.
The resolution concluded that the EU's AI strategy must not overlook the military and security considerations and concerns that arise from the global deployment of AI technologies. Members stressed the challenge of reaching a consensus within the global community on minimum standards for the responsible use of AI and expressed concern about military research and development on autonomous lethal weapons systems.