The Power of Defaults: Google’s Payout Playbook Threatens OpenAI’s Dominance

Google cornered the search market by weaponizing defaults, and Judge Mehta cleared the way for this to be extended to AI, forcing OpenAI into a hardware gamble

Around 12 years ago, I was out with grad school friends in San Francisco when we witnessed something that stuck with me.

A woman at the bar was upset about her relationship. When my friend asked why she was with him, her answer was simple: money. Her boyfriend’s father was a high-ranking tech executive. In fact, he is one of the richest people alive.

The dynamic was uncomfortable to watch, but it illustrated something important: money complicates everything.

Look no further than Google’s patronage program, where they pay browsers like Apple’s Safari and Mozilla’s Firefox for making Google their default search engine.

Google has already extended that patronage program for Gemini, making it the default LLM on Samsung devices, and that money could really complicate OpenAI’s distribution strategy.

Google’s Patronage: The Monopoly that Must Go On

In 2024, US District Judge Mehta ruled Google’s patronage program violated Section 2 of the Sherman Act, as their massive annual payments kept search competitors out of the market.

• Apple: $20 billion

• Samsung: $2 billion

• Mozilla: $555 million

Their payments were so big, in fact, that in 2025 Judge Mehta later ruled that these illegal payments could continue! He admitted that a ban on Google payments for search placement would cause downstream effects, specifically:

• “Lost competition and innovation from small developers in the browser market”

• “Less investment in the U.S. market by Android OEMs, which would reduce competition in the U.S. mobile phone market with Apple.”

• “Higher mobile phone prices and less innovative phone features”

And now that Judge Mehta has allowed these payments the obvious question is: what if Google followed the same strategy so Gemini is the default LLM on all browsers and devices?

That’s what would keep me up at night if I worked for OpenAI. Defaults are as powerful as they are underrated, and Gemini has a chance at cornering their consumer AI distribution by following the same playbook as search.

Apple Maps and the Power of Defaults

Did you know that more than 20% of Apple’s $99 billion net income is an annual payout from Google? Google pays Apple over $20 billion each year so Google search is the default on Safari.

Defaults are powerful, and Google learned this lesson painfully in 2012 when they overplayed their hand with Apple in Maps. Google planned on defeaturing their iOS Maps product compared to the Android version, while demanding more user data and more Google branding on the Apple version.

Apple was furious, and released a competing product in Apple Maps, going down in history as one of the worst product launches ever. It was famously lampooned in the HBO show Silicon Valley.

And despite being one of the worst launches ever, Apple Maps still took meaningful share away from Google. That’s the power of defaults, as Google testified in their own antitrust case (bold emphasized by me).

Two years after Apple Inc. dropped Google Maps as its default service on iPhones in favor of its own app, Google had regained only 40% of the mobile traffic it used to have on its mapping service, a Google executive testified in the antitrust trial against the Alphabet Inc. company.

Michael Roszak, Google’s vice president for finance, said Tuesday that the company used the Apple Maps switch as “a data point” when modeling what might happen if the iPhone maker replaced Google’s search engine as the default on Apple’s Safari browser.

Google ensured this error would never happen again, and ramped up their patronage program, where they pay browsers and devices for making Google their default search engine.

And Google’s market share has been dominant ever since.

The beauty of Google’s payouts is that competitors increasingly can’t get a foothold. Google’s revenue keeps increasing, which lets them pay browsers and device manufacturers even more money.

And this makes search even better, as all of those users create a positive feedback loop. You can see why this violated the Sherman Act.

But after Judge Mehta’s ruling, why wouldn’t Google extend their payouts, so Gemini is the default LLM as well?

Massive Financial Disparities Between OpenAI & Google

Google should pay for making Gemini the default, because OpenAI can’t possibly match them, just follow the money.

There are a few big takeaways from this.

• Google makes 148% more net income than OpenAI has ever raised each year.

  • $116B net income vs $78B raised by OpenAI

• 20% of Apple’s net income comes from Google alone

And if that 20% from Google is very consequential, so would any incremental dollars. Apple’s net income has been stagnant over the past 2 years. Their highest YoY growth has been 4% over that time, and their largest income drop was -8% (blue line).

Factor in that OpenAI’s 2025 Burn Rate is $8 billion, and any Apple payout would accelerate their burn rate.

Conversely, Google could match OpenAI’s total funding, and still have a healthy net income that year. Looking at this financial reality, why wouldn’t you do this if you were Google? And given Apple’s stagnant income, why wouldn’t they take the money?

Implications for AI Competition & OpenAI’s Response

An Apple & Google deal would be a big issue for OpenAI, as Gemini is already embedded across Android devices, which have over 70% global market share.

Now, Judge Mehta has opened the door for Google to escalate their payments to Apple, cornering AI in mobile just as they did in search.

OpenAI has one critical strength, which is their 800 million weekly active users. They still have commanding market share in the consumer space.

This seems like the steelman argument for Apple choosing OpenAI. Apple could brand themselves as the “ChatGPT phone”, and seize some of Android’s users who would prefer a native ChatGPT experience.

But how likely is that? Apple is notorious for owning the user experience, remember the Apple Maps story we mentioned? Giving up that much of the user experience would go against everything Apple is.

And there could be a question about ChatGPT’s durability. Yes they have 800 million weekly active users, but Google Search has 4 billion users, and Gemini is landing number 1 in the App Store.

It’s also possible that Gemini becomes superior, as speculators are betting Google has 70% odds of having the best AI model by end of 2025.

Looking at this again if you’re Apple. Is it worth giving up less money from Google when they have 4 billion search users, are climbing up the App Store, and are surpassing OpenAI in LLM capabilities?

OpenAI’s Response: Vertically Integrate with Hardware

It seems like OpenAI is aware of these risks, as they’re proactively acquiring companies for their own hardware distribution.

In May 2025, they acquired io Products, which was co-founded by Apple’s legendary designer Jony Ive. They also announced a partnership with Luxshare, a key Apple supplier.

Recent rumors suggest OpenAI is developing a smart speaker, glasses, voice recorder, and a pin.

If these products take off, OpenAI can bypass their reliance on Apple & Android. But that is a big if, as consumer hardware is notoriously tough. Most products end up flopping, like the Essential Phone, Amazon’s Fire Phone, and the Humane Pin.

But the best time for displacing a new market is when a new paradigm arrives. Of course, Apple & Android won the mobile market from Nokia, Motorola, and Blackberry thanks to the smartphone paradigm. We’re now in a new paradigm with generative AI.

And just like that relationship at Smuggler’s Cove, Apple and Google may not have a perfect partnership, but they know what to expect. Given that, an Apple deal with Gemini feels imminent. So OpenAI might as well seize the new paradigm and hope their hardware finds traction.

Apple Maps showed how powerful defaults can be when they took market share away from Google Maps, and Google’s Gemini is a far better product than Apple Maps.