EU Court of Auditors: EU must do more for AI
Outdated and vague investment targets, lack of coordination and monitoring of investments in the EU are criticized by the auditors.
Emblem of the EU Court of Auditors
(Image: eca.europa.eu)
The European Union has so far had little success in developing an artificial intelligence ecosystem in Europe. This is the opinion of the EU Court of Auditors in a special report (PDF). The EU has failed to accelerate AI investments in order to catch up with the world's leading companies.
The EU Commission has done a lot since 2018 to advance the EU's AI ecosystem. By this, the Court of Auditors means areas such as regulation, infrastructure, research and investment. Steps were also taken early on in the EU to investigate AI risks, which led to the world's first general rules for the use of AI with the AI Act. However, the steps were not well coordinated with those of the member states and investments were not systematically monitored. In future, more - and more targeted - public and private investment is needed if the EU wants to achieve its AI ambitions.
USA in the lead
The EU Court of Auditors defines AI as new technologies in areas such as robotics, big data and cloud computing, high-performance computing, photonics and neuroscience. The USA has long been the leader in these areas, while China plans to become the global AI leader by 2030. In the years 2018 to 2020, the planned private investments in the EU amounted to 20 billion euros and 20 billion euros in each subsequent year, the Court of Auditors calculated.
In the EU, the proportion of companies using AI varies considerably between the member states. According to the Court of Auditors, France and Germany have announced the largest public investments in AI, while four countries still have no AI strategies at all.
Overall, venture capital investment in the EU has been lower than in the other AI-leading regions of the US and China since 2015, according to the Court of Auditors. It is estimated that the overall investment gap between the US and the EU more than doubled between 2018 and 2020. The EU has therefore taken steps to develop a coordination framework for AI by expanding investment and adapting regulation. In 2018 and 2021, the Commission and the member states agreed on measures to develop an AI ecosystem.
"Investment targets vague and outdated"
The Commission's AI plans for 2018 and 2021 are in line with international best practice, writes the EU Court of Auditors. However, the EU's investment targets are too vague and outdated and have not changed since 2018. "The lack of ambition for investment targets is at odds with the goal of building a globally competitive AI ecosystem." Although the EU Commission has increased EU budget spending on AI research projects, it has not significantly increased private co-funding. The Commission must also do more to ensure that the results of EU-funded AI research projects are fully commercialized or used.
(Image: Europäischer Rechnungshof auf der Grundlage der jüngsten Eurostat-Daten)
More than five years after the EU's first plan, the framework for coordinating and regulating investments in AI is still being worked on. So far, the EU Commission's coordination with the member states has only had a "limited impact". "This was because the executive lacked the necessary governance tools and information", according to the EU auditors. The EU Commission had not set up a suitable system to monitor AI investments. There is also no overview of how the member states are contributing to the EU's general investment goals.
(anw)
