Something's got to give: European regulators crack down on AI as U.S. tech companies plunge into the market

By Jon Swartz

Tech companies are watching their actions in the more progressive E.U., where regulations are years ahead of those in the U.S.

An ever-expanding assault on Big Tech by the European Union has U.S. companies scrambling to adhere to rules there while trying to not slow their dash to embrace generative artificial intelligence.

The contrasting timelines have U.S. tech companies watching their actions in the more progressive E.U., where regulations are years ahead of those in the U.S.

The E.U. AI Act, which is the world's first major set of regulatory ground rules for governing AI, has been in development for three years. The legislation puts controls around AI use cases anywhere. "Europe is NOW a global standard-setter in AI," Thierry Breton, the European commissioner for internal market, wrote in a post on X, formerly known as Twitter.

As part of the enforcement, which is expected to go into effect between May and July, the E.U. summoned a dozen large tech companies - including Microsoft Corp. (MSFT), Facebook parent Meta Platforms Inc. (META), Alphabet Inc.'s Google (GOOGL) (GOOG), Snap Inc. (SNAP), TikTok and X - to disclose how they are using AI.

"It boiled down to: 'Are you being honest and transparent?'" Asha Palmer, senior vice president of compliance at software-as-a-service company Skillsoft, said in an interview. "This is very similar to the GDPR-related risks. The relationship between the government and private industry is a bit of a dance, a tango. Who is leading whom?"

Palmer was referring to the E.U.'s General Data Protection Regulation, which went into effect in 2018.

"For regulations to work, there needs to be enforcement," she said.

In March, the E.U. launched a probe into Big Tech's use of artificial intelligence and its handling of computer-generated deepfakes, ramping up scrutiny of a technology that officials fear could be used to disrupt elections. Regulators at the European Commission say they're particularly concerned about how generative AI could sow chaos in the run-up to this summer's E.U. parliamentary elections.

Read more: E.U. opens probes of Apple, Alphabet, Meta under new law

The investigation is under the auspices of the recently enacted Digital Markets Act, which is designed to rein in the largest tech companies - the watch list includes Alphabet, Amazon.com Inc. (AMZN), Apple Inc. (AAPL), Meta, Microsoft and TikTok parent Bytedance - that act as gatekeepers between consumers and other businesses.

E.U. lawmakers have adopted the motto "Don't break up big tech companies - break them open."

In Europe, there is growing skepticism about the reach of tech policies. Regulators are focusing on rules and legal frameworks instead of creating an environment for innovation, some argue.

"The E.U. is not really driven by any market logic with all these regulations," Luigi Congedo, an entrepreneur and former venture capitalist, said in an interview. "Unfortunately, I am afraid entrepreneurs are not supported, and most of the regulations are actually just blocking entrepreneurial initiatives instead of supporting them."

DMA's far-reaching impact

Above all else, the specter of the Digital Markets Act has elicited uncharacteristic alarm among tech giants. Apple has claimed the DMA jeopardizes the security of its devices, while Google has warned of "difficult trade-offs" for search results. Meta, Apple and TikTok have all filed legal challenges against the E.U., claiming that the new laws unfairly target their services.

Among these tech behemoths, the fear is over what the DMA may lead to: iPhone customers would be able to download apps from platforms other than Apple's App Store. Microsoft Windows would no longer have Microsoft-owned Bing as its default search tool. And Google and Amazon would have to tweak their search results to create more room for rivals.

"The nature of AI is that it is completely dependent on data ... and getting better outcomes," Ameesh Divatia, the CEO of data-protection company Baffle Inc., said in an interview. "It is not good to be late to gen-AI development. It is better to be in the market, and then change as the law dictates."

European tech companies have been through this before with the E.U.'s General Data Protection Regulation, and they have geared up for the DMA. There are no European companies on the DMA watch list - only U.S. companies and the Chinese TikTok.

SAP (SAP) has an advantage, largely because it is headquartered in Germany and has dealt with GDPR rules since they were implemented in 2018. The software company, which has major operations throughout the U.S., has also been a late and cautious adopter of generative AI.

"As a European company, we are much more on the edge on responsible AI use," Yaad Oren, who heads the innovation labs at SAP's (SAP) campus in Palo Alto, Calif., said in an interview. "We believe in the triple R's of AI: relevant, responsible, reliable. AI can be like shooting a fly with a cannon."

Data protection and risk assessment is a "a particularly European thing," said Efrain Ruh, chief technology officer for Digitate, a software company with operations in U.S., Europe and Asia. "There is a lot of hype with genAI, and many companies are jumping in too fast. We have been cautious in approach and risk."

"In some ways, we are seeing the next development in the use and regulation of Big Data," said Paul Lekas, senior vice president of global public policy at Software & Information Industry Association, which has nearly 400 members including Amazon, Google, Meta, and Apple.

"The EU AI Act has changed quite a lot, and there is a lot of confusion and uncertainty over how provisions are implemented," Lekas said.

Some U.S. tech companies have also been trailblazers in helping craft responsible AI use worldwide.

Adobe Inc. (ADBE) has engaged regularly with policy makers to establish guardrails for the responsible development and use of AI.

Adobe launched the Content Authenticity Initiative - an association promoting the use of labels to show whether content is AI-generated or not - in 2020, with the New York Times and Twitter (now X) as the only other members. Since then, the initiative has swelled to more than 2,500 members and is helping the Coalition for Content Provenance and Authenticity, known as C2PA, develop an industry-standard label to show where a piece of content comes from and how it was created.

Still, the DMA undoubtedly will disrupt the speed of what many U.S. companies hope to do with generative AI in Europe. But that could play to the advantage of late bloomers such as Apple, which can develop technology that adheres to new laws.

"Remember, OpenAI was late and small and [was] still capable of becoming a leader," Appian Corp. (APPN) CEO Matt Calkins said in an interview. "It's almost better not to be first, and learn from others' mistakes."

-Jon Swartz

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

04-06-24 0634ET

Copyright (c) 2024 Dow Jones & Company, Inc.