The European Parliament on Tuesday adopted a report on artificial intelligence, which sets out a list of demands to secure the EU’s position in AI, and points to research as one of the key means to achieving that goal.
MEPs warn the EU must move quickly to set clear rules for AI if it wants to have a say in the future of the technology.
“We have the opportunity to set global standards,” said Parliament’s rapporteur for the file, Axel Voss, speaking in the final plenary debate. “If we allow ourselves to lose leadership position, we will resign ourselves to the status of digital colonies subjugated to other regions that don’t share our values.”
The report is the culmination of a year and a half of work by the Parliament’s special committee on AI. It will feed into work on the upcoming AI Act, the first major AI regulation globally, which will set rules for AI uses according to their level of risk.
As things stand, the US is leading the AI race in terms of investment, research and attracting talent while its companies spearhead technology development. China is behind the US, but catching up.
The EU has a long way to go to, but policymakers believe with clear regulations and an investment push it can catch up with competitors.
The aim is to increase public and private investment to €20 billion by the end of the decade. Today, it stands at €14 billion.
While Parliament wants to see more money flowing to research, it says clear rules may need to come first.
“The number one barrier certainly is market fragmentation, which in turn affects investment and research. Without a truly harmonised digital single market, the resulting lack of cross-border investment and also cross-border data exchange prohibit innovation of any kind,” Voss told Science|Business.
Parliament’s vision for fixing the issue, as set out in the report, starts with increased EU investment in AI and other key technologies, supported by a strategic roadmap.
Parliament wants member states to boost investment too, both within their national budgets and through strengthening the EU research and digital programmes, Horizon Europe and Digital Europe.
To make life easier for researchers, the Parliament urges the European Commission “to simplify the structure of research funding, including grant application requirements and processes.” This should go hand in hand with improving the quality and consistency of reviews of proposals to ensure the right projects get funded at the right time.
On education, Parliament wants to see more teaching posts on AI at universities that come with adequate salaries.
There also needs to be supporting infrastructure, and MEPs urge the Commission to establish AI data centres and stress the importance of testing sites.
Cooperation between institutions will be key to establishing excellence hubs that can attract talent. For this, the Parliament calls on the Commission to establish AI lighthouses as part of Horizon Europe, to form networks of research organisations that are involved in AI.
“Lighthouses or even lighthouse regions are a way to pool and combine European resources toward enabling innovation across borders and testing new products and ideas, in a manner outside of fragmented rules and procedures,” said Voss.
The Commission is already working on one network of AI excellence centres. Margrethe Vestager, Commission vice president responsible for digital policy said this aims, “to bring together excellence, because we know that talent attracts talent, and if we want the best to come here, we need to invest.”
All these efforts should be focused on sustainable and socially responsible AI, in line with EU values.
The AI Act will follow on the heels of other EU regulations in the digital sphere that the Commission has been churning out over the past months. These include the Digital Services Act and Digital Markets Act aiming to curb the influence of big tech, which were adopted by policymakers in recent weeks.
The AI Act is expected to be negotiated early next year but negotiations between the three EU institutions are yet to start.