The EU cannot shape the future of AI with regulation

The European Commission recently announced plans to increase that funding, to make more data available for use in AI, and to work with EU member states on a strategy for deploying AI in the European economy.

But at the same time, the EU's new General Data Protection Regulation (GDPR) puts tight restrictions on uses of AI that involve personal data, and EU policymakers continue to search for additional restrictions on AI to address their remaining fears.

Unlike tobacco, AI has many beneficial uses, and the potential risks depend on how it is developed and used over the long-term.

The irony is that if Europe over-regulates AI now, it will miss its chance for global influence over the technology's future.

The commission does not see a contradiction because it believes that stringent regulation will engender consumer trust in AI. But that reasoning is flawed.

Regulation can help to build confidence in a technology when there is a specific problem with a clear solution. But like a drug, regulation has side effects, is not appropriate in every case, and overdoses of it can be devastating.

AI is still in its early stages, both technically and economically. AI systems can perform a narrow range of tasks quite well, but training algorithms to do anything especially new is difficult, costly, and time-consuming.

The practical uses of AI that exist in the real economy are valuable and encouraging, but they remain peripheral to the activities of most businesses.

AI implications unclear

That means two things.

First, the implications of advanced AI are not yet clear, so neither are the policy measures that would properly address risks.

Second, if the EU adopts more regulations that raise the costs and legal difficulty of using AI at such an early stage, adoption of the technology by firms in Europe will fall even further behind the rest of the world.

That, in turn, will limit Europe's capacity to innovate in practical applications of AI, reduce European competitiveness, and diminish the EU's relevance as a global decision-maker in the long-term future of AI.

To be sure, the commission's recently-published package of data policies will help to make more data available for use in AI.

The package included a proposal to strengthen the Public Sector Information (PSI) Directive, which requires European governments to allow third parties to reuse the data they publish, though the amendments still would not address the core problem of governments not publishing enough data in the first place.

It also included soft-law measures to push academics to make more of their research data available, and to encourage businesses to find ways of sharing more of the valuable data they hold.

More data is good, and so is the commission's intention to boost funding for AI research and to work on a comprehensive strategy with member states. But over-regulation can kill even promising technologies that otherwise have strong government support.

If the EU wants to shape the future of AI, then it should give firms and inventors in Europe the best chance of building and deploying cutting-edge AI systems that the rest of the world will want to adopt.

That will give the EU a much louder voice when it comes to answering questions about safety and ethics, and it will give international partners a reason to work with the EU in finding the right answers.

But if the EU continues to over-regulate AI, its AI systems will fail to compete on a global scale and the technology's long-term future, for better or worse, will be shaped by the United States and China.

Nick Wallace is a Brussels-based senior policy analyst at the Center for Data Innovation. His Brussels Bytes column deals with the digital single market and data-related policy issues in the European Union