The European Union’s new AI and data strategies come with a hefty vein of ethical positioning running through them and a barely hidden threat to ‘tame’ US tech giants that don’t live up to Europe’s self-proclaimed high standards.
The overall ambition of the bloc is clearly articulated:
Harnessing the capacity of the EU to invest in next generation technologies and infrastructures, as well as in digital competences like data literacy, will increase Europe’s technological sovereignty in key enabling technologies and infrastructures for the data economy.
What’s lacking is much detail on how such high-minded soundbites and mission statements will translate into specific action and what impact such action will have in terms of (a) encouraging (again) the emergence of European tech contenders and (b) the impact on investment in EU states by non-EU tech leaders.
In its European Strategy for Data issued by the European Commission yesterday, the proposal is made to establish a “single European data space” to allow digital information to flow freely. To encourage that and to force larger firms to share data with smaller rivals, the Commission says that it may make access to such data “compulsory… under fair, transparent, reasonable, proportionate and/or non-discriminatory conditions”.
Thierry Breton, Commissioner for the Internal Market, commented:
Our society is generating a huge wave of industrial and public data, which will transform the way we produce, consume and live. I want European businesses and our many SMEs to access this data and create value for Europeans – including by developing Artificial Intelligence applications.
On AI, the White Paper for Artificial Intelligence: a European approach to excellence and trust talks about enforcing ethical behavior through regulation, including rules to ‘train’ AI systems using data that is representative of the total population to tackle conscious and unconscious bias; requiring owners of AI systems to keep records of how they were developed; making it compulsory to alert citizens that they are interacting with AI tech rather than human beings; and re-training AI systems developed outside the EU so that they comply with European rules.
The Commission proposals recognise that not all AI is created equal. So there is ‘high risk’ AI, in use in sectors such as healthcare, transport and justice, that needs to regulated strongly and be demonstrably transparent and subject to human oversight and scrutiny. High-risk applications, including those from non-EU providers, would have to be certified compliant with EU standards before being used in member states. Then there is ‘lower risk’ AI tech for which a lighter touch of voluntary labelling and ‘trustworthy AI’ certification will suffice.
Despite predictions and leaks from Brussels to the contrary, the Commission’s proposals steer clear of an outright ban on facial recognition technology. Instead the proposals recommend an ongoing debate about what exceptional circumstances could permit the use of such systems.
The proposals are now open for public consultation and comment until 19 May, so expect a flurry of corporate lobbying to get underway. Draft legislation could follow by the end of the year and be signed into force in 2021.
Some laudable aims, some instinctive knee jerk moves to be seen to get tough with US “Big Tech’ and a lot of wooly wording that doesn’t really get down to nitty-gritty on how this vision will be delivered.
There’s also a certain lack of ambition that is frankly disappointing. For example, the proposals want to pump €20 billion a year into AI over the next decade, which simply isn’t going to put the EU on a level investment playing field with the US or China.
Despite this, there’s another airing for a familiar battle cry from the Commission as Breton declares:
The battle for industrial data starts now and Europe will be the main battlefield. Europe has everything it needs to be a leader.
All of this rhetoric - which we’ve heard many times before on tech policy matters and to little long term effect - is influenced by the wider ideals of the ‘European Project’ with European Commission (EC) president Ursula von der Leyen talking about:
Creating incentives for data sharing, establishing practical, fair and clear rules on data access and use, which comply with European values and rights…the capability that Europe must have to make its own choices, based on its own values, respecting its own rules. This is what will help make tech optimists of us all.
That’s admirable enough as a value statement, but in practical terms what is the impact on innovation of fresh layers of regulation? There’s a fine line between protecting citizens rights and taking an ethical stance and smothering organizations with red tape.
The ‘laissez faire’ approach of the US to data is one extreme, while the centralist control of China is another. The danger that the EU needs to avoid is creating a third extreme of aspirational envy of the US tech sector combined with regulatory and protectionist instincts that don’t provide the necessary incentives to stimulate growth.
The EU is the leader in regulation, not innovation and that’s a problem. Commissioner Breton made no secret of the ‘catch-up’ aspect of the new proposals:
We recognize that we missed the first wave, or the first battle, which was the battle of personal data.
For its part, the US Government is (perhaps predictably) unimpressed by what it sees coming out of Brussels. US Chief Technology Officer Michael Kratsios said in a statement:
We encourage the EU to follow America's lead and pursue an innovation friendly, values-based approach to AI regulation, one which avoids over-burdensome, one-size-fits-all policies. The best way to counter authoritarian uses of AI is to ensure the US and its allies remain leaders in innovation, advancing technology underpinned by our common values.
After a big build-up, what emerged yesterday from Brussels was distinctly underwhelming. That said, the two white papers are proposal documents intended to kick off a public debate. The path to turning the ideas contained within to EU law is a long and complicated one, so there is - theoretically - time to put some more meat on the bones.
As an aside, what stance post-Brexit Britain will take on whatever emerges remains to be seen. With the UK government’s current public position being that it will not be forced into automatically aligning with EU standards, the debate over the next few months may well end up as another complicating factor in negotiating trade deals with both the EU and the US.