Conquering generative artificial intelligence (AI) is the next battlefield for big tech platforms, but Google could be on the back foot thanks to its dominance in search.
The proposed remedies recommended by the Department of Justice (DOJ) to rid Google of its alleged monopolistic behavior in search have the potential to hamper its future developments in gen AI.
Google, responding to the multiple potential remedies from the DOJ, naturally wants to pull focus back to the $20 billion deals it struck with Apple to keep its search on rival devices, rather than entertain some of these border topics, like data sharing with rivals and splitting up its crawlers.
"Even though data powers Google's AI products, Google is not light years ahead in the AI race at this point, and, arguably, is a couple of steps behind Microsoft and OpenAI," said Alan Chapell, president of privacy-focused law firm Chapell & Associates. "And search, as we knew it, like Googling information, has already evaporated."
While it's not known the ultimate direction Judge Amit Mehta will take, what is clear is that it will be a long ride. Proposed remedies won't take place until April of next year, with a ruling in August, which Google plans to appeal.
Whatever the decision, there's a delicate balance between keeping Google's power in check and promoting a more competitive landscape, sources told ADWEEK. There will also be the matter of preserving innovation and keeping in tact systems that the rest of the industry, especially smaller players, rely on.
One proposal from the DOJ includes "requiring Google to allow websites crawled for Google search to opt out of training or appearing in any Google-owned artificial-intelligence product or feature."
Publishers have groused about being unable to opt out of Google's crawlers that scrape their sites for gen AI use cases and model training, while still being able to be indexed in search.
Sure, publishers would need to weigh up not appearing in Google Overview, and the potential traffic or revenue that could generate, but that would also free them up to strike deals with other AI tech firms like OpenAI, Perplexity, or Anthropic to sell the data the platforms are hungry for.
"For a publisher the economics are favorable," said Forrester senior analyst Mo Allibhai. "Each AI can come to them. It's all incremental. It could give them more leverage."
Granted, it hasn't always felt like publishers, especially smaller ones, have had much leverage in this gen AI arms race. However, OpenAI has struck deals with publishers like WSJ worth more than $250 million over five years, meanwhile Google reportedly pays Reddit $60 million per year to gain access to its data.
Another potential remedy from the DOJ is requiring Google to share aggregated, anonymized search data with rivals like Microsoft and DuckDuckGo to help them build their search businesses.
"Splitting off or divesting Chrome or YouTube from Google would be less catastrophic for Google than forcing it to share its privately shared information," said Erik Hamilton, vp search & social at agency Good Apple.
Sharing that data comes with privacy risks, but perhaps not enough that it's dismissed by a judge, writes Casey Newton, author of Platformer newsletter. Plus, we're already seeing something similar in action in Europe. Last year, the European Union's (EU) Digital Markets Act took effect to put some data-sharing requirements on "gatekeeper platforms" like Google. Now search engines that serve European users can apply to get Google search data for more than 1 billion queries across 30 countries.
Jeremy Hull, chief product officer at agency Brainlabs, notes that there are two types of data the DOJ is asking for Google to make available.
The first are Google search results, features, and ads, including the underlying ranking signals, especially on mobile. Having access to those precious underlying ranking signals, which for over a decade have, and continue to, confound marketers, publishers, and competitors opens up the opportunity for rivals.
The second type of data relates to the indexes, data, feeds, and models used for Google search, including those used in AI-assisted search features. This, according to Hull, would have a bigger immediate impact on marketers.
That's also a double-edged sword. Marketers might be keen to get their hands on some of that data Google is prompted to share with rivals, but it will increase the complexity of management and testing of day-to-day operations dramatically, he added. That's no good for marketers, especially SMBs, with tight resources.
Performance Max, which uses gen AI to run ads online, relies on data tracking across the web.
The outcomes of any of the global adtech antitrust lawsuits could seriously hamper Google's ability to gather the data that makes PMax so effective. That could be a critical weakness.
"If that goes away, I wonder how good these audience signals will be," said Sam Tomlinson, executive vice president at boutique agency Warschawski.
He added: "You take away all these different data points to probabilistically model out an audience you're back to a very keyword-heavy, very manual very limiting ecosystem for a lot of small and midsize advertisers."