By Walter van der Weiden
For some time now, Europe’s progress on Artificial Intelligence (AI) has compared unfavorably to developments in the US and China. The finger of blame has pointed in familiar directions: fragmentation of private and public R&D, lack of scale up capacity and funding, a significant number of European SMEs have yet to adopt digital solutions, general inability to tap into a continent-wide data pool and loss of local talent. But there is growing optimism that this is about to change, and a new sense that Europe knows where it wants to go on AI.
The EU and some leading member states are taking steps to ensure Europe catches up on AI. For example, the recent European Commission AI Communication, as well as comprehensive policy plans and funding (to be) proposed by France, Finland, Germany and the UK amongst others, offer plenty of food for thought. These initiatives are indeed steps in the right direction. In fact, the underlying good news is that there appears to be general consensus that Europe should fully embrace AI – but mindful always that the human focus should be central to its AI strategy. Thus, the European AI angle from a policy perspective focuses more on enhancing social welfare and consumer trust, facilitating technology that better complements skilled workers, and staying within well aligned ethical boundaries, with active enforcement where necessary.
Important however will be to steer clear from self-defeating protectionist measures and avoid “winner takes all” outcomes which create further inequality, power asymmetries and bias. Additionally, AI innovation should not be at the expense of consumer welfare and basic protections; building the necessary trust with the wider community is crucial for long term policy success, effective implementation and consumer acceptance. Following the experience with GDPR which might become the de-facto global data protection rulebook, Europe could in fact lead global quality standards for AI on algorithm transparency, avoidance of bias, liability, data quality, and consumer protection. This is an opportunity for EU to take a strong stance in terms of ethical leadership when it comes to AI and may be a defining moment in how this nascent industry moves forward. Whilst it is certainly too late to resist AI developments (especially with US and China aggressively investing in it), Europe can play a more meaningful role in shaping the future direction of AI and ensuring that innovation should not be at the expense of consumer protection and well-being. As a start, it is important for Europe to recognize and focus on the following 3 key priorities:
- AI must be broken down into its separate components to identify which specific areas Europe should lead on, reap the most important benefits from, and can most likely contribute to long term growth and economic prosperity. AI is not one single issue, and it should not be simplified. In some areas, for example, Europe already has well established industries (e.g. automotive sector). In other areas like e-commerce, Europe is significantly behind other regions. However, it is clear that Europe’s strength lies in its manufacturing and industrial base, which should weigh in on policy makers’ decisions;
- Regulatory ‘sand boxes’ for AI developers, should become an EU-wide initiative.This recent excellent French proposal, if extended to the whole EU, wouldpromote early development in Europe of AI technology, services and startups, through setting up a targeted light-touch regulatory model;
- Data access across EU borders should become the EU Internal Market’s 5th Freedom. Being the indispensable “fuel” for successful AI, access to massive amounts of quality data will be essential. EU-wide access to quality data would put Europe on a more equal footing with China and the US, whilst at the same time ensuring regulatory protections are in place.