MEPs debate the future of AI with Margrethe Vestager and Thierry Breton
MEPs from the special committee on Artificial Intelligence in a Digital Age (AIDA) discussed ways to ensure that AI applications meet EU standards and values, on Monday and Tuesday.
You can watch the recordings of the debates here.
“I am glad that Executive Vice President Margrethe Vestager and Commissioner Thierry Breton joined us for AIDA’s first two official hearings, and that they are willing to continue this dialogue throughout the mandate of the committee.” said AIDA committee Chair Dragoş Tudorache.
“Through the establishment of AIDA, Parliament has recognised the subject of Artificial Intelligence as of strategic interest. Its establishment responds to the need to have a place where we can address the larger, umbrella issues related to Artificial Intelligence, to analyse its future impact on a number of key sectors, to evaluate third countries’ approach to AI, and to provide a strategic plan with medium-and long-term objectives for AI in Europe” he said.
“AIDA members come from all relevant committees and from all political groups. It is the first time that such strategic subjects can be addressed, taking into account different competencies, within a coordinated Parliament voice”.
“Topics related to AI need to be approached from multiple and reinforcing angles: AI diplomacy, exporting EU ethical standards at the global level, European strategic resilience, and maintaining EU competitiveness in the face of global advances in AI – they cannot be taken piecemeal”.
“AIDA thus complements the work of Parliament’s standing committees, without overstepping on their competences. We want to bring concrete added value to Parliament’s work and to our citizens” he added.
During the debates, MEPs raised questions on access and control over the data classification foreseen under the upcoming legislative proposal on data governance (the so-called "Data Act") to be tabled in 2021, and on ways to help start-ups and SMEs get an easy access to the future European data market. MEPs also spoke in favour of ensuring that the widespread use of AI does not lead to unintended or disproportionate citizen's surveillance. Other MEPs questioned the Commission on ways to reduce the technology's carbon footprint, help the reskilling of workers, as well as on the relevance of the taxation of AI applications and robots.
"Ecosystem of trust"
Discussing the EC’s White paper on Artificial intelligence with MEPs, Executive Vice-President Margrethe Vestager spoke in favour of developing "a human-centric AI that is rooted in the values of European society" and vowed to create "an ecosystem of trust" with clear, predictable rules for AI that are of high-risk to breach safety or fundamental rights. For low-risk applications, the approach should be light-touch, she explained. "All this potential will not be unleashed if we cannot trust the technology" and identify reliably the different stakeholders, she said.
On the future “Data Act”, Internal Market commissioner Thierry Breton said that “there is no AI without data, so we have to get ready for the incoming wave of industrial data, which will be even bigger than the wave of personal data” he said. “Europe is an industrial continent, and has the potential to be a leader” in this area, he added. The quality of the future datasets will be key to ensure the development of EU-based AI services. The challenge will be to ensure that data generated in Europe remains under European jurisdiction while keeping access to future European data market open to third-country actors playing by European rules. Future AI services might have to go through a risk-based conformity assessment before they can be placed on the EU market, he said. Standardization and ammonization of datasets will be two critical factors to ease interoperability and trustworthiness he added. Regarding algorithms, “transparency” and “explainability” are key requirements for high-risk applications that have a heightened risk of fundamental rights breach, such a bias and discrimination.
In the same meeting, the AIDA members had the chance to discuss OECD’s work on artificial intelligence with Director for Science, Technology and Innovation, Mr Andrew Wyckoff, and Deputy Director, Mr Dirk Pilat.
Background
The European Commission released a white paper on Artifical intelligence in February 2020, opening a reflection and a public consultation ahead of future legislation to be tabled in 2021. Making Europe fit for the digital age is among the priority policies of the EU executive for the legislature. The European Commission is also to propose in December a Digital Services Act, on which the European Parliament gave its input in a resolution adopted on 20 October.