Qualcomm Was Found Guilty of Monopoly. Then a Higher Court Erased Every Word of It.

On May 21, 2019, a federal judge wrote the two words every monopolist dreads: 'anticompetitive malice.' Judge Lucy Koh found Qualcomm guilty on all three theories the government had thrown at it, and ordered it to tear up its licensing model and renegotiate contracts worldwide.³ It was, by any reading of the headlines, a corporate execution. Fifteen months later a higher court read the same record and reached the opposite verdict — unanimously, and on every single point.⁴ Same company, same conduct, same evidence. Two governments split against themselves over what it meant.

The story everyone remembers is that Qualcomm got caught running an illegal monopoly, that Apple was the brave whistleblower, and that Apple won. Almost every beat of that is wrong. Qualcomm's model survived the law. The DOJ sided with Qualcomm against its own sister agency. And Apple, the loudest accuser, ended up writing Qualcomm a check for billions before the trial it had demanded ever started.

The toll nobody could route around

To understand the war you have to understand the move at the center of it: 'no license, no chips.' Qualcomm sold modem chips, the baseband processors that connect a phone to a cellular network. But it also held a thicket of patents on the cellular standards themselves. Its rule was that you could not buy the chip unless you also took the patent license — and the license was priced as a percentage of the whole phone, not the chip. The FTC called this monopoly maintenance and sued on January 17, 2017, alleging Qualcomm refused to license rival chipmakers and locked Apple into de facto exclusivity.¹ Three days later, on January 20, Apple filed its own suit.² The sequence matters: the regulator went first, and Apple followed into the opening.

Here is the thesis, and it cuts against the headlines: Qualcomm's model was hyper-aggressive but not antitrust-illegal. The Ninth Circuit's reasoning was almost clinical. Qualcomm had no antitrust duty to license its patents to rival chipmakers. Its chip policy, however harsh, did not harm competition in the market the case actually defined. And disputes over whether its royalties were 'fair, reasonable and non-discriminatory' were contract and patent questions, not Sherman Act questions.⁴ Koh had been right that the conduct was harsh. She had been wrong about which body of law governs harsh. That distinction — between unfair and illegal — is the whole case.

When the government argued against the government

The strangest character in this drama is the United States, because it showed up on both sides. The FTC had brought the case and won the first round. Then the Department of Justice — another arm of the same federal government — filed an amicus brief siding with Qualcomm, warning that Koh's ruling could undermine American leadership in 5G just as the race with China sharpened.⁸ This is not how antitrust enforcement usually looks. The agencies present a united front; here they were openly at war. And the appeals court noticed: the Ninth Circuit found the policy arguments persuasive and folded that logic into its reversal.⁸ Qualcomm's deepest defense was never a clever contract clause. It was that breaking its model meant kneecapping the one American company carrying 5G against Huawei — and a court of appeals agreed that mattered.

Apple's surrender, and who actually paid

Now the part that rewrites the popular memory entirely. On April 16, 2019 — with a jury trial about to begin — Apple folded. The two companies dropped all global litigation, and Apple paid Qualcomm.⁶ Not the other way around. Qualcomm's own SEC filing estimated $4.5 to $4.7 billion in settlement revenue, a blended figure combining Apple's payment with the release of Qualcomm's obligations to refund Apple and its contract manufacturers.⁵ The settlement also handed Qualcomm a six-year global patent license effective April 1, 2019, with a two-year extension option, plus a multiyear chipset supply agreement.⁶ Apple, in other words, agreed to keep buying Qualcomm's chips and keep paying Qualcomm's royalties — the exact arrangement it had sued to escape.

Why would Apple capitulate weeks before the trial it had spent two years building? Because it was cornered by physics, not by law. Apple had moved its modem sourcing to Intel, but Intel could not deliver a competitive 5G modem on Apple's timeline. To ship a 5G iPhone, Apple needed Qualcomm — and Qualcomm knew it. The settlement was not a verdict on the merits; it was a 5G dependency forcing Apple's hand at the worst possible moment. The accuser settled not because it was losing in court, but because it could not afford to win and still build a phone.

But didn't Qualcomm really exclude rivals?

The honest objection is that this reads too kindly. Qualcomm did pursue exclusivity with Apple — the FTC's own press release alleged it precluded Apple from buying competitors' baseband chips from 2011 to 2016, a five-year window in which no rival could get in.⁷ And there is a real argument that Koh saw the commercial reality clearly and the Ninth Circuit hid behind a narrow market definition. Fair. But two things hold the thesis up. First, the law is the law: an appeals court does not reverse unanimously on a record this scrutinized unless the legal theory was genuinely overreached, and the DOJ's break from the FTC tells you the overreach was visible from inside the government.⁸ Second, the proof is in the settlement. If Qualcomm's position were as weak as the 2019 ruling suggested, Apple — armed with the world's best lawyers and a regulator already winning — would not have paid Qualcomm billions to make it stop.⁶ Capitulation by the accuser is its own kind of verdict.

Qualcomm spent three years as the textbook villain — found malicious, enjoined worldwide, sued by both a regulator and its largest customer at once. And it walked out the other side with its model intact, its largest accuser back on its payroll, and a federal appeals court holding that aggressive is not the same as illegal. The crisis it survived was never really about the chips. It was about which court, which statute, and which definition of harm gets the last word. Qualcomm lost every headline and won the only sentence that counted.