The Executive Branch of the European Commission is working hard to protect Europeans from a promising future.
By proposing to regulate artificial intelligence, the rapidly changing technology that is fueling much of the tech revolution, the EC is building walls that will stymie innovation in Europe, and constrain leading information technology firms in the United States as well.
AI is a catch-all phrase for computer software that gathers information, looks for patterns in the data, and then helps businesses, governments and individuals make better decisions. AI is instrumental in how Amazon recommends books. An international team is using AI to track the coronavirus. AI drives Tesla cars and helps spot fake news on Facebook.
The potential of AI is enormous. Eighty-five percent of CEOs worldwide believe that AI will significantly affect how they do business. PricewaterhouseCoopers estimates that AI will add nearly 15% to the US gross domestic product in 10 years.
But as with any new technology, AI has a dark side. Self-driving cars have caused fatalities, but so do human drivers. AI used in the justice system has exhibited racial bias, but that was because the underlying data included racial bias. AI sometimes makes wrong medical diagnoses, but doctors often make better diagnoses when assisted by AI.
No matter how promising a new technology is, it can trigger fears. Often governments respond to these fears with new laws that promise protection, but that include their own dark side that impedes a promising future. This is what the EC has done.
The EC’s proposed policies would place mandatory constraints on high-risk AI applications, such as health care, transport, policing, and the judiciary. They would create laws to impose safety controls and new legal liabilities on future AI developments. And impose restrictions on the use of information about Europeans, beyond those recently adopted as the General Data Protection Regulation.
The EC claims that its regulations will help European companies become more innovative and provide world leadership on AI regulation. But those claims ignore the evidence of how such regulations affect an economy.
Europe’s GDPR has proven to be a business killer since it was implemented in 2018. The regulations impose new obligations on companies that gather and use large amounts of data. Research at the Illinois Institute of Technology and the National Bureau of Economic Research found that GDPR has led to a more than 25 percent reduction in the number venture capital deals in Europe compared to the United States. Research has also shown that the more regulated an industry, the harder it is for startups to overtake incumbents.
The EC policies will repeat these experiences by limiting what AI can do in Europe. The limitations on data will raise the costs of trying to develop AI innovations. Startups will have long waits while risk-averse bureaucrats study possible problems. And the restrictions on what can be done in medicine, transportation, and other “high-risk” areas will stop valuable applications: The most important data in these contexts is that which is gathered when humans intervene to keep things from going wrong. The EC’s attempts to keep AI safe will also keep it from being as effective as it can be.
The EC hopes that it will be a world leader in AI regulation. Let’s hope not. Non-EC countries would be better off following the lead of President Donald Trump, whose executive order on AI emphasizes supporting research and development, and keeping regulation out of the way so that consumers and businesses can discover which AI innovations are valuable and which are not.
EC laws and regulations on AI have implications for the United States. Europe has spent considerable time investigating U.S. tech companies and collecting billions of dollars in fines. In addition to draining companies’ resources, the rules will have a chilling effect on new investments. New U.S. companies will be cautious about entering European markets. And the regulations will hinder innovations that U.S. companies could create if they had better access to European markets.
It seems unlikely to happen, but the EC would do well to follow the United States’ lead on AI. A knowledge economy runs on learning, and the EC’s current direction “protects” businesses from doing that. Europeans and Americans would both benefit from more innovative and vibrant businesses from Europe.
Dr. Mark Jamison is the director and Gunter Professor of the Public Utility Research Center at the University of Florida, and also directs the digital markets initiative and serves as a visiting scholar with the American Enterprise Institute.
Morning Consult welcomes op-ed submissions on policy, politics and business strategy in our coverage areas. Updated submission guidelines can be found here.