Chrome is the Antitrust Problem, Selling It is Not the Solution

Google is worse than Microsoft ever was. With a stranglehold on search and online advertising, backed with an Orwellian surveillance of users, dominating the browser market is too much. The solution cannot be to sell Chrome to OpenAI, however.

Often, attempts to fight monopolies are actually consumer antagonistic in result. Breaking up large companies is easy; making things better is not.

I’ve written about the Department of Justice’s ill-advised attempts to fight easy, secure messaging, for example. Likewise, Europe’s insistence on new devices presenting “choice” screens concerning browsers and app stores is the opposite of helpful to novice users. How does an average person make an informed decision? Why make it difficult to get the option provided by the brand they actually chose to buy a device from?

The best antitrust enforcement gets out of the way and lets better alternatives win by their own merits. Microsoft’s late-1990s run-in with antitrust authorities is a good model.

The company had a lock on desktop operating systems and leveraged that to the advantage of its office suite and web browser. Microsoft penalized PC makers that considered alternatives and actively introduced code for the purpose of breaking competitors’ software.

In the aftermath of the antitrust action, the then-gatekeeper to 90% of the computer market had to compete on merit. The company’s loss of monopoly control ultimately boiled down to a lack of will to make the best products.

The early 2000s saw Microsoft Internet Explorer stagnate into a buggy mess. IE was so bad that when Mozilla released Firefox, average, non-techie users went and downloaded that breath of fresh air upstart.

The open source Mozilla, accompanied by the launch of Apple’s Safari, also featuring an open core by way of KDE, changed the market. Firefox never took the majority of the market, but became enough of a presence that websites could no longer lazily amplify Microsoft’s control by writing bad code that only worked on Internet Explorer.

Firefox jumpstarted Microsoft’s web browser development again, but it was too late. Google’s Chrome quickly supplanted Firefox, but either way, users were now accustomed to the idea that the internet was not Internet Explorer.

Meanwhile, Apple’s iPod seed grew into the iPhone. Given an opening by Microsoft CEO Steve Balmer’s startlingly bad foresight on where mobile computing was headed, the iPhone with its Safari-powered web browser — and rejection of Adobe Flash — pushed us towards a more open web that worked on any device.

Google’s Android and Chrome both drew on the foundation of WebKit, Apple’s open source code behind Safari. With WebKit on every popular mobile device and a significant force on the desktop, the Web was no longer hostage to the stagnant and buggy.

Chrome emerged right at the inflection point when Google was transitioning from the “Don’t Be Evil” white knight to the new villain of tech. For the most part, Chrome was another step forward in those early years, finishing the job Firefox started of unseating the aged tyrant, Internet Explorer.

Alongside Google and Apple, once significant players Nokia, BlackBerry and Palm also all came behind Apple’s WebKit open source project. The free, open web appeared to win the day.

All besides Apple and Google faded, however. As Google’s dominance solidified, the company ultimately “forked” that code, parting ways with Apple’s effort and dousing the brief hope of a unifying, vendor-neutral web browser engine. After that split, Chrome’s base remained open, but it was directed by Google for the benefit of Google.

While Microsoft’s control of the internet was the browser itself, so an inferior browser made them vulnerable, Google’s control has always been what is inside the browser. The company spun its search monopoly into a new era of web browser dominance. Microsoft eventually made the “if you can’t beat ‘em, join ‘em” move of rebuilding its own Edge browser on Google’s core code, but even then, most users only open Edge in order to dutifully download Chrome.

Chrome is very much to our present time what Microsoft Internet Explorer was to, say, 2004 — bloated and inferior — but users pass up superior built-in browsers (Edge or Safari) and install Chrome, anyway. It’s Google, after all, and thus works the best with the company’s monopoly in search and dominance in other realms such as maps, e-mail and basic office documents. As in the bad old days, some sites refuse to work with alternatives.

The DOJ is thus right to see Google’s web browser dominance at the center of an antitrust resolution. But, a blunt force antitrust approach to Chrome, spinning it off entirely, lays the foundation for new antitrust abuses.

Take OpenAI’s interest in buying Chrome. Divining Sam Altman’s motive for guiding his juggernaut to make the purchase is not hard. On a surface level, it would give OpenAI direct access to customers for ChatGPT subscriptions. Under the surface, they could — and almost certainly would — record what we do and what we view. OpenAI’s business depends on devouring as much data as possible to stay ahead in the ever-accelerating AI arms race; who believes it could resist the siren call of such a treasure trove?

Google’s long-term interests in Chrome have been to push its other products. To be charitable, Google wanted Chrome to look the other way as its owner tracked our information for the sake of its advertising business. Chrome has consistently lagged behind browsers such as Safari and Firefox in protecting our privacy.

But if OpenAI buys Chrome so GPT can watch us and learn more about what it means to act “human,” is that in any way better?

Even if OpenAI only tracked us as much as Google does now, the company is already threatening to supplant Google as people favor AI over traditional search. The DOJ pressuring one monopoly into creating a possibly worse future one is reckless, not responsible.

We benefit if American (and other Western) AI companies like Altman’s do well, lest the People’s Republic of China gain a perilous lead, but only within limits. Altman and OpenAI are not well suited to being tasked with returning us to the “Don’t Be Evil” era.

As the queen in Google’s chess game, simply selling Chrome to a company like OpenAI solves nothing. It would merely change which company controls the side of the board putting us into checkmate.

If the government really wants a consumer-friendly solution to Google’s dominance, Chrome’s history should again become its future. Instead of wresting the Chrome browser away from Google, emancipate the underlying Chromium Projects, the open source underpinnings that power it.

Most of Chrome is already open source code within Chromium, but Google tightly controls Chromium and users seldom see Chromium until it is “Googlified” into the final Chrome product. Passing this browser “engine” off to an independent project outside of Google could provide equal opportunity for queening other browsers’ pawns.

Admittedly, there would be challenges. Chromium as an independent project might suffer from the less complete symbiosis with its former parent. Or, if Google continued to be the dominant contributor, a different organizational structure might do little to change the company’s de facto anticompetitive power.

But, what if the DOJ leaned on Chrome’s historical origins and brokered Chromium’s reunification with Apple’s WebKit Project? Reversing the twelve-year-old split would lay the foundation for a genuine “reference implementation” browser. Cousins WebKit and Chromium remain the base of every significant non-Mozilla browser; merged together, “ChromeKit” would be the standard-bearer.

Ostensibly, the W3C leads web standards, but those are all theoretical in the form of papers laying out specifications. Some see the light of day, others do not. ChromeKit could implement new standards that would directly and quickly feed into Chrome, Safari, Edge, Opera and beyond.

A single browser, even a single open source one, would not be ideal. Competition is necessary to avoid another era like the early 2000s’ near-complete stagnation. This is not that, but rather a healthier headwaters all the downstream competitors could both be beneficial to and benefit from.

Operated meritocratically like many open source projects, such as Linux, those who contribute would direct its future. With enough of Big Tech involved, no one player would be powerful enough to sway it to their exclusive benefit.

Given that only Google is under DOJ pressure, could a ChromeKit become reality? There’s every reason to think so.

Apple’s interest is in having the best browser to show off their hardware, so this would be almost pure benefit to them, no doubt why they welcomed Google as a contributor to WebKit prior to the fateful 2013 divorce. Apple would gain immediate fruit from Google (and Microsoft) contributed code, lessening their own costs in implementing technology that shows up in Chromium first.

Meanwhile, Microsoft, too, has nothing but upside. In ChromeKit, the company would be one of several peer contributors to the code its Edge browser depends on, rather than a vassal.

Even for Google, this is the best possible antitrust resolution, since the company could still personalize the final Chrome product to its branding and vision. Much as Microsoft plowed along post-antitrust two decades ago, Google could keep producing its products. But, user-hostile privacy choices, only sensible for the sake of lining its advertising pockets, would be curtailed.

A ChromeKit Foundation would likely need to steward both WebKit and Chromium for a time. Diverging developments laid down over a decade are not trivial to bring together. But, the end payoff would justify the effort. Getting to that end could also be accelerated by any cash component of the final antitrust enforcement funding the work.

Every antitrust action should aim to improve things for end users, not just break things. If Google and the DOJ could agree to antitrust relief in this untraditional, but technologically aware, form, it would be the opposite of those user-hostile European “browser choice” screens.

Browsers’ less brand-specific components wouldn’t be implemented and reimplemented across different cores. New standards could appear in final, consumer-ready browsers faster. And, whichever browser a user chose, even lazy web developers’ pages would display properly, since pages would be displayed consistently from browser to browser. Alternatives could thrive without the cost of inferior user experience.

WebKit as such a centerpiece of an open web was nearly a reality a decade ago. The DOJ has a generational opportunity to try revivifying that dream for us all. Will it take it?