Sharp Tech (preview): The App Store in the shadow of AI

Summary

• Ben Thompson’s meta-takeaway on Apple’s regulatory changes is that nobody cares anymore—including him. The FT reports Apple “acknowledged for the first time” that regulatory loosening is weighing on its $100B+ services business, alongside an EU DMA settlement, Apple’s proposal of 15% commissions for standard apps in the United States, and a German ruling requiring the same ATT consent-request language for Apple’s offerings and third-party apps. The man who once obsessed over the App Store mock-falls asleep: “And now no one cares. I think that’s the bigger point. Including me.”
• AI shifted attention to a new growth story, and growth makes zero-sum fights feel less important. “Obsessing about the App Store is emblematic of stagnation”: when the pie grows, “all the problems that inevitably exist… don’t really matter that much”; when it’s fixed, slicing it becomes “overwhelming, all-encompassing.” That’s “what the EU gets wrong by and large by undervaluing growth.”
• The same attention shift hits TikTok—even though Ben says it may matter more than ever. “The TikTok issue is actually the biggest issue facing tech right now”: is China’s thumb on the scale pushing content to discourage certain AI data-center buildouts? Ben calls that a “reasonable concern” and says China would be “derelict in their duty” otherwise; Andrew agrees, while acknowledging the issue commands less passion than it did three or four years ago.
• The under-covered App Store story: Apple’s constraints on app-building services. For months, Apple was not letting apps made through Replit-like services update themselves unless they stripped out functionality, including the ability to run other apps. Ben grants “legitimate security concerns” and praises the iPhone’s sandbox, but says Apple “wants a cut of everything.” Shortcuts are a powerful but inscrutable AI use case, and “Apple is choking off something of the future” while the issue feels less pressing.
• Germany’s ATT ruling is the right outcome, four years too late. ATT is “hilariously anti-competitive, and anti-competitive in a very classic sense”—scary prompts for rivals, soothing wording for Apple’s own tracking, and rules changed after the fact in Apple’s favor. Ben’s verdict: “This is a good outcome. They were thorough”—but the ad ecosystem was already remade in the interim.
• The regulatory timing lesson: better to catch the bad actor late than kill the growth engine early. “The App Store was and is awesome”—a foundation for trillions of dollars in value and “astronomical consumer surplus”—and “if you worry about the bad things too early… the good things don’t happen… you regulated nothing.” Andrew’s coda: he is not as anti-regulation, but AI is changing so quickly that it’s “almost guaranteed we will get more wrong than right” if regulated before we know how it will be used and how it will work in practice.

Deep dive

1. Apple concedes on the App Store—and the ex-obsessive falls asleep

• Andrew’s setup, via the FT: Apple has “acknowledged for the first time” that regulatory changes forcing looser App Store control are weighing on its more-than-$100B services business. He also cites an EU settlement over the DMA, Apple’s proposal of 15% commissions for standard apps in the United States, and Germany’s competition authority holding that Apple can no longer use different consent-request wording for its own offerings and third-party apps. Ben’s response: “Oh, sorry, I fell asleep because the topic felt so boring.”
• Ben was early—Felix Salmon once cited the App Store as his classic newsletter “pet obsession no one cares about”—and the fight reached a fever pitch during COVID, when companies moved in-person classes online and Apple demanded 30%. That period nearly burned him out: he still enjoyed tech and content production, though getting clipped was annoying, but five-hour congressional hearings and regulatory work were what depressed him about the job.
• Two or three weeks after writing “Aggregation Theory,” he predicted regulation and court cases were inevitable and argued that applying antitrust law to aggregators would not work.

2. Growth is the political solvent; stagnation is the fight over slices

• The through-line: App Store obsession was “emblematic of stagnation”—people fighting to re-slice a pie that stopped growing. “This is what the EU gets wrong by and large by undervaluing growth.”
• On Epic: he appreciates Tim Sweeney “fighting the fight” philosophically, but a multibillion-dollar corporation “arguing over who gets a bigger percentage of us selling virtual dances”—“purely zero marginal cost”—earns little sympathy. “And now no one cares… Including me.”

3. TikTok may matter more than ever, and Apple is choking app building

• Andrew’s admission: he once would have written “several 3,000- or 4,000-word pieces” on TikTok; now he would struggle to muster the same passion. Ben’s flip—worth keeping—is that “the TikTok issue is actually the biggest issue facing tech right now”: is China pushing content to stop certain AI data-center buildouts? Ben calls that a “reasonable concern” and says China would be “derelict in their duty” otherwise; Andrew agrees.
• The under-covered App Store story concerns services such as Replit that let people make apps. For months, Apple was not letting these apps update themselves unless they stripped out functionality, including the ability to run other apps. Ben grants “legitimate security concerns” and praises the iPhone sandbox—built from day one to isolate apps—but says Apple wants everything to go through the App Store for a cut. He points to Shortcuts as a powerful but hard-to-use AI use case: AI could help people build them. “Apple is choking off something of the future,” but the issue stays quiet because it is not pressing while everyone feels opportunity.

4. Germany’s ATT ruling: right call, four years late

• ATT is, in Ben’s view, “hilariously anti-competitive… in a very classic sense”: Apple imposed scary tracking prompts on competitors while its own asked, “Oh, we want to give you a better experience”—for “the exact same thing for all intents and purposes.” Ben adds that restrictions honestly disclosed on the tin are not necessarily antitrust violations under U.S. law; Google’s problem in the Epic case was changing the rules after the fact. Apple, by contrast, changed the rules after the fact while favoring itself.
• Ben’s archive moment: searching for his old link, he found that his “Data and Definitions” article opened with this very German investigation launching—“I didn’t even realize it was the same case… it’s been four years.” Verdict: “They were thorough, and I think the right thing is happening”—while acknowledging the frustration that the advertising ecosystem had already been remade around ATT.

5. Better to regulate the backside late than never get the front end

• Ben ties the EU AI Act to the EU having relatively little going on in AI and to the Act’s limitations. He says the impetus was partly the feeling that “we passed the DMA too late”; the Act is now driving a watermarking discussion. He characterizes Mistral’s big play as “basically” being a data-center operator for Chinese models.
• The closing argument: the App Store “was and is awesome,” laying the foundation for trillions of dollars in value and astronomical consumer surplus. Ben would rather have “all this upside with a bit of Apple being a bad actor on the back end” treated too late than obsess over the back end early—“in which case you never even got to the back end, so you regulated nothing.”
• Andrew’s AI coda: he is not as anti-regulation, but AI is changing so quickly that it’s “almost guaranteed we will get more wrong than right” if regulation comes before we know exactly how it is used and how it will work in practice.