This week the European Commission will release a draft of proposed policies outlining how the European Union will grapple with the rapid spread of Artificial Intelligence in just about every aspect of our lives. As the New York Times reports, these proposed regulations promise to be far-reaching enough to prompt Alphabet’s Sundar Pichai and Mark Zuckerberg of Facebook to travel to Brussels to meet with Margarethe Vestager, the Vice-President of the European Commission tasked with developing AI policy. The exact scope of the proposed regulations should be clearer when the Commission releases its report on February 19. However, Vestager has already called for “particularly high standards regarding transparency and accountability” in implementing AI.
The “European Values” Driving Artificial Intelligence Policies
The urgency with which the European Commission is formulating an EU-wide AI policy shows the importance that issues of technology and privacy are viewed within the European framework. When the current President of the Commission – the former German Minister of Defense, Ursula van der Leyen – took office on December 1 last year, she announced that Brussels will formulate a comprehensive AI policy in 100 days (in addition to separately formulating policies dealing with climate change and minimum wage levels in each EU member.)
But the debate over the ethical implications of AI has been simmering in the EU for years. For example, in June 2018 the European Economic and Social Committee (EESC) held a “Stakeholder Summit” on Artificial Intelligence to discuss how Brussels can help guide the impact of AI on spheres such as legal and ethical challenges, social and economic impacts, and economic competitiveness. The EESC had also called for “a standardization system for verifying, validating and monitoring AI systems,” including a “uniform global code of ethics for the development, application and use of AI.” These statements point to a greater likelihood that creators of AI technology will have to contend with a singular (and comprehensive) Europe-wide system rather than a patchwork of policies depending on which country decides to tackle the issue, and how far-reaching that country is willing to regulate AI.
German Skepticism Toward AI
But another source of resistance in the EU for creators of AI technology just might be the consumers themselves. A recent survey asking Germans about their attitudes toward AI showed widespread resistance to smart-home technology, with 57 percent of respondents saying that would not want to live in a home in which AI assists people—even if that technology were free of cost. The reasons for such resistance included fears that personal data could be used illicitly, potential dangers of becoming too dependent on the technology, and a wish to avoid feeling “powerless” in the face of AI. An overwhelming majority of respondents – 85 percent – stated that products utilizing AI should be clearly marked and inspected by a independent authorities.
Brussels Will Outpace Washington in Setting Regulatory Frameworks for AI
While a debate about the ethical implementation of AI has by no means been absent from the United States—particularly discussion about how AI could either reflect or deepen already-entrenched biases—the issue has only recently been gaining traction in Washington. One effort to address the ethics of AI has been the Algorithmic Accountability Act introduced in 2019 aimed at avoiding disparate treatment and discrimination. But this bill would not in itself set guidelines for how AI should operate ethically but rather require the Federal Trade Commission to prepare such rules that would apply to companies developing AI. Given the current makeup of Congress the Act seems unlikely to be passed anytime soon, and even if it were the FTC would then need to go through the rule-making process (the Act requires the FTC to create such rules within two years.) Brussels on the other hand appears ready to act even if this week’s report might be less comprehensive than anticipated or does not take effect until a later date. And whereas the Algorithmic Accountability Act focuses mostly on combating bias and racial discrimination, the European Commission has been discussing AI in a broader ethical framework that encompasses more wide-ranging concerns about human dignity. It therefore seems likely that AI companies will need to consider the requirements of Brussels—especially given predominantly German concerns over technology and privacy—while seeing very little pushback (if any) from the U.S. Congress.
GermanLegalBlog.com 