What does the Google antitrust ruling mean for the future of AI?

When the government launched an antitrust lawsuit against Google five years ago, it was all about whether the tech giant had a monopoly on internet search. But by the time the judge ruled on punishments for the company this month, the future of artificial intelligence was front and center.

Federal district court judge Amit Mehta tried to thread a needle with his latest ruling, rapping Google for its stranglehold over search without hobbling its ability to compete in the newer field of generative artificial intelligence as chatbots grow in importance.

Mehta recognized the challenge, writing, “the court is asked to gaze into a crystal ball and look to the future. Not exactly a judge’s forte.”

Some analysts fear he may not have done enough to put Google in check, though, because the vast trove of data its search engine collects gives the company a huge advantage when it comes to developing AI. They also say the case highlights the limits of the judiciary when it comes to crafting policy for fast-moving technologies.

The Department of Justice had sought stiff penalties, including Google’s divestment from its Chrome browser, a corporate crown jewel, and a ban on exclusive deals the company made with device makers for Google to be their default search engine.

But Google gets to keep Chrome, and financial deals to make Google the main search engine can happen — just not exclusive ones. The company will also have to share some of its valuable search data with competitors, although those details haven’t been nailed down yet.

“I think when you look at the package of the remedies all together, they ring hollow,” said Alissa Cooper, an internet policy expert and executive director of the Knight-Georgetown Institute, a think tank focused on tech policy. “It doesn’t really propose very significant interventions of any kind.”

Chatbots, like Google’s Gemini and its rivals, ChatGPT, Claude and Perplexity, are now nipping at the heels of traditional search, as people increasingly use them to answer questions instead of typing queries into search engines like Google.

Cooper says the judge seemed to see AI as a completely new battlefield that will have its own set of winners and losers. And that’s why he didn’t lay down strict punishments that would prevent Google from turning its strengths in search into strengths in AI.

She said he seems to believe “we should trust that the competition that emerges from these new players will take care of the problem that led to bringing this case in the first place.”

Like a parole officer?

Even with steps the judge took to try to rein in Google’s dominance of internet search, Tim Wu, a Biden administration tech advisor and professor at the Columbia University law school, says it’s not clear if this antitrust case will stop Google from throwing its weight around to dominate AI.

“The question is, as far as I’m concerned, does Google get to use the same techniques that it used to win and dominate search — old search — for AI?” he said.

Judge Mehta’s decision to bar Google from making exclusive distribution deals was worded in a way that it applies to AI chatbots, sometimes called “answer engines,” as well. He also assigned a technical committee to oversee the firm in the coming years, and ostensibly make sure it plays by the rules. Wu said it’s unclear what impact that panel will have.

“Is this like a parole officer who Google has to check in with every, you know, month or so? Or is the court sort of done with it?” he said.

Either way, he and others argue, the government should be doing more to regulate tech policy than just filing court cases.

Tom Wheeler, a former chairman of the Federal Communications Commission, agrees. He said this case shows that courts can’t keep up with technology, particularly in a case that hinges on the Sherman Anti-Trust Act, which is more than a century old.

“You’ve got to feel for Judge Mehta because he was trying to apply a law written in 1890 to what’s going on today in artificial intelligence,” he said.

Wheeler says antitrust cases are insufficient for regulating a fast-evolving market like AI. He notes that competitiveness in the AI race will be fueled by data, reach, computing power, and money. “The reality is that Google controls all of those,” Wheeler said.

In court this spring, Google’s attorneys argued that when it comes to AI, the company is not top dog. The pacesetter is OpenAI’s ChatGPT. Google’s Gemini chatbot is a competitor, and Google gives brief AI answers to some search queries.

Google declined to comment for this story, but has said it’s likely to appeal both the penalties and Mehta’s underlying verdict that it has a monopoly in search.

“This case will ultimately be determined by the Supreme Court, probably in the 2027, 2028 timeframe, which is seven or eight years after it was filed — and a millennium in terms of innovation and development,” Wheeler said.

Note: Google is a financial supporter of NPR.