Nieman Foundation at Harvard
Google’s on trial in Washington, D.C. Not metaphorically, either: A federal court is almost halfway through what’s expected to be 10 weeks of testimony on whether Google’s search dominance is a violation of U.S. antitrust law. The primary issue involves Google’s multi-billion-dollar deals with browser makers — the ones that make Google their users’ default search engine — and whether they are an unfair barrier to competition. (According to the Justice Department, about half of all Google searches come from a browser the company pays to be the default — and about half of all the rest come from Chrome, which Google owns.)
The trial thus far has given us some remarkable looks inside the strategic thinking of some normally opaque tech giants. On Monday, we had Microsoft CEO Satya Nadella acknowledging his company’s Bing is no match for Google in quality. “Defaults are the only thing that matter,” Nadella testified, “in terms of changing user behavior.”
The most important of those defaults is set by one company: Apple. Nadella testified that he’d offered Apple up to $15 billion a year to switch Safari’s default to Bing — even saying he’d be willing to hide all Bing branding and let Apple brand the search engine as its own.
But Google pays Apple plenty to head off that possibility. (The size of Google’s current annual payment to Apple is not public — both companies claim it’s a trade secret — but reports have suggested it’s north of $15 billion, perhaps approaching $20 billion.) All of this has unfolded amid reporting that Apple has invested huge amounts of money into building its own search engine tech — some of which is used inside current Apple apps but which could evolve into a full-blown search engine.
How will Judge Amit Mehta rule once all the testimony is in? Who knows? But there is, for the first time in a long while, a distant twinkling of a possibility that Google’s search dominance could take a meaningful dent. (Mehta could force Google to stop buying browser default status with all those billions. Or to sell off its dominant Chrome browser. Or to allow other search engines access to its mountain of search data. Or, of course, he could say Google’s dominance is perfectly fine and change nothing.)
This is all important stuff which, as Tim Wu writes, “will effectively establish the rules governing tech competition for the next decade, including the battle over commercialized artificial intelligence, as well as newer technologies we cannot yet envision.”
But — to be parochial for a moment — what sort of an impact would any movement in this direction have on the news business? That’ll depend on the remedy, obviously. But Google search is by far the No. 1 source of external traffic to news sites, and any meaningful competition for that title could shake the foundations of the industry. To be clear, we’re in thought-experiment territory: Google will be the dominant search engine for the near-to-medium future and any real shift is at least years away. But here are a few of the possibilities.
Apple is in an extraordinary and unique position when it comes to search, needing only to flip a switch in Safari to become the biggest non-Google search company of the past two decades. Its iPhones make up a majority of all smartphones in the United States and large shares of most European countries, along with small-to-medium shares around the world. Around 35% of all web browsing (mobile and desktop) in the U.S. and 20% in Europe and worldwide happens in Safari. By making Sherlock (or whatever Apple would call it) the default search engine in Safari, a large share of those users would be immediately switched into it.
Would some switch back to Google? Of course — some, maybe even most. When Apple decided to build its own Maps app to replace Google Maps back in 2012, reviews were terrible — but over time, Apple Maps has become a strong competitor. That’s the worse-case scenario: If Apple ever decides to launch a search engine, I’d wager it’ll be a much better Day 1 product than Maps was. And while Google can offer integrations between search and its other products (Gmail, Gcal, Google Maps, and so on), Apple can offer a deeper list of its own, since it controls the entire operating system and device.
That, to me, is the most likely path to real search competition in the United States. Apple’s users are unusually attractive to advertisers because they are, on average, more affluent, and more interested in buying things. (And more interested in reading news online.) So even if an Apple Sherlock didn’t become No. 1 in market share, it would punch above its weight.
The key reality for individual publishers vis à vis Google has long been that they have zero power in the relationship. Sure, they could decide to remove their site from Google’s search index if they don’t like the grand content-for-traffic bargain the company offers. But in the real world, pulling out of Google only hurts the publisher. Google will be fine without The Daily Gazette’s stories; The Daily Gazette won’t be fine without Google’s traffic.
Would that change if there were two big search engines in town? The internet user in me hates the idea of the web being split among multiple search engines — certain results being a Google exclusive or Apple exclusive. But both would no doubt try to differentiate themselves from each other. Think of how Apple Music has invested in exclusive albums, live performances, and programming to distinguish itself from Spotify. Or how Spotify spent more than $1 billion on podcasts to distinguish itself from Apple Music. Those strategies sometimes work and sometimes don’t — but they do often involve shoveling money at the people making the content.
Would an Apple search engine, say, consider paying a small fraction of a search’s ad revenue to the site that a user clicks onto — the site that, in effect, won the user’s attention? Google would never consider such an idea because, well, it has no incentive to. (Keeping all the money works fine for them!) But if a new player (with $167 billion in cash sitting around) thought it would gain market share, could something like that sound interesting?
That may sound far-fetched — heck, it likely is far-fetched! — but this is a unique moment in the relationship between the tech giants and publishers because of two letters: AI.
Building a search engine involves spidering the entire internet — finding every webpage you can and copying their contents into a massive search index, all so some later user can type “what was matlock’s first name” into their browser and be told it’s Ben. American courts have held that all this ingesting and copying is nonetheless legal under copyright law because it counts as a “transformative” fair use.
A few decades later, along comes AI. Like a search engine, building a large-language model involves ingesting an enormous amount of stuff in order for the AI model to “learn” the stuff’s contents. Does that sort of massive copying count also count as a “transformative” fair use?
The courts haven’t decided that one yet. But it seems the main AI companies are getting ahead of the problem — agreeing, in concept, that some publishers should be able to sell the right to use their content in AI training. Some deals have already been struck.
Satya Nadella referenced this in his testimony. Here’s The Verge’s David Pierce:
While Nadella said AI has the potential to shake up the market a bit, he also believes it could further entrench Google’s dominance. Search engines are fundamentally reliant on websites allowing themselves to be crawled so that the engines can understand their content and direct users to those sites. Nadella called search engines “the organizing layer of the internet” in that sense.But as large language model-powered systems continue to grow, publishers and platforms are growing wary of how their data is ingested and used to train AI systems. Those publishers, Nadella said, may start to sign exclusive deals that allow Google — and only Google — to use their data. That would essentially crush every other search engine on the market.
“What is publicly available today, will it be publicly available tomorrow?” he asked; “That’s the issue.” He said he’s already hearing from publishers with a deal in place with Google, asking Bing to match it.
Because of the legal fuzziness around AI, negotiations over platform uses of publisher content are a thing. But as Nadella notes, it risks creating a world in which the strongest players only get stronger, raising a huge barrier to entry.
Still, it doesn’t take an MBA to understand that it’s better to have multiple tech giants angling for your content than to have just one. Google is happy to hand out money where it thinks it serves their interests, but it’s unlikely to rethink its operating fundamentals without an outside force — like the government and/or Apple — intervening.
This potential universe — where Apple and Google split the search market, much as they already do the smartphone OS market — is hardly the stuff of Bernie Sanders fantasies. (“So you’re telling me the big new competitor for a $2 trillion company is…a $3 trillion company?”) There’s no reason to be pollyannaish about Apple as a partner to publishers; it’s just as profit-hungry as Google and it lacks Google’s commitment to/reliance on the open web — always happy to build its own walled gardens. The ideal solution, of course, would be lots of competition, from the biggest giants to the tiniest startups — which is something Judge Mehta could be in a position to catalyze.
My point isn’t that Apple would be a better tech overlord. It’s that competition is a good thing, even if it’s coming from two of the richest corporations in the history of humanity. Ideas move from unthinkable to, well, thinkable. Search is a critical — for many, the critical — supplier of audience to news publishers, and the rules for how it works are frozen because of one player’s market dominance. The groundwork can be laid now for those thousand flowers to bloom — but I suspect it’ll take another giant pushing Google off its pedestal first.
