European legislators have proposed new laws that could put strict limits on AI. The legislation is moving forward even though a recent study finds the regulations could cost the EU economy around $36 billion over the next five years. Some observers argue that such measures are needed in the US, as well. “As our society transitions towards a digitally enabled environment, an unregulated AI could lead to misuse, potentially undermining our rights to privacy and data protection,” Joseph Nwankpa, a professor of information systems at Miami University in Ohio, told Lifewire in an email interview. “In addition, poorly regulated AI could reinforce inherent societal biases by exploiting imbalances of information access across various segments of our society.”
Cracking Down on AI
AI poses many threats, experts say. One growing area of concern is the use of AI to spread misinformation through deep fakes, Wael AbdAlmageed, a professor at the University of Southern California who studies AI and other subjects, told Lifewire in an email interview. It also could lead to insurance fraud by manipulating evidence and endanger child safety by using AI to generate fake identities. Europe is ahead of the US when it comes to AI legislation. The Artificial Intelligence Act (AIA) is a proposed law recently put forward by the European Commission, the executive arm of the European Union. It’s not just European countries that would be affected under the EU law. The AIA would apply to any provider of AI whose services or products reach the EU market. Laws would regulate AI tools used in everything from financial services to toys. The act bans AI systems that use subliminal techniques to “manipulate a person’s behavior in a manner that may cause psychological or physical harm.” It also prohibits “exploiting vulnerabilities of any group of people due to their age, physical, or mental disability in a manner that may cause psychological or physical harm.” AI would not be allowed to provide real-time remote biometric identification in publicly accessible spaces by law enforcement, except in specific public safety scenarios. US lawmakers are moving to rein in AI, as well. The Algorithmic Justice and Online Platform Transparency Act of 2021 aims to prohibit the discriminatory use of personal information by online platforms and require transparency in algorithmic processes, Nwankpa said. “Like many other types of technology, it’s not the technology itself that will be regulated, but the manner in which companies and individuals use the technology,” Dara Tarkowski, a lawyer who specializes in the intersection of law, technology, and highly regulated industry, told Lifewire in an email interview. “Certain uses of AI are already regulated. Remember Apple Cards? Regulators are already examining the impact of AI on fair credit and lending, for example.”
Debate Rages Over Regulation
Not everyone agrees that artificial intelligence should be regulated. AbdAlmageed said that governments shouldn’t pass laws that control how AI is used and developed. “However, I believe AI scientists and engineers should exercise some form of self-censorship in terms of what algorithms they develop, how they evaluate it, and how they deploy these algorithms in real-life products,” he added. AI is especially challenging from a regulation and technology standpoint, Jason Corso, director of the Stevens Institute for Artificial Intelligence, told Lifewire in an email interview. “We can see automobiles; we can see seatbelts,” he pointed out. “AI primarily operates behind the scenes. It is data; it is software; we cannot see it. To make the matter worse, many products are being advertised as ‘AI’ when they are not, which is partially a problem of AI being ill-defined and partially a problem of over-zealous marketing.” Despite the risks posed by AI, legislation to regulate the technology faces strong opposition in the US. Tarkowski said she doesn’t think that the US will pass laws as broad as those under consideration in Europe. But, she said, “I anticipate US regulators will incorporate guidance and possible amendments to existing laws where AI may have an impact—such as the Equal Credit Opportunity Act.”
