Ai coding flood chokes apple's human app store reviewers
Apple’s pride and joy—its human-curated App Store—is drowning. A tsunami of AI-generated apps, spun up in minutes by “vibe coders” who barely know Swift from salsa, has stretched the once 24-hour review window into a week-long limbo. Developers now watch their release calendars melt while Cupertino’s flesh-and-blood reviewers, famed for saying no, suddenly can’t say anything at all.
The manual gate meets the machine gun
Inside Apple Park, the math is brutal. Every 24 hours, the queue swells with thousands of new binaries whose core logic was hallucinated by large language models. Humans still open each one, click through every screen, hunt for hidden crypto miners, check for copy-cat icons, test against 30-page guideline PDFs. The average review time has tripled since February; outliers top 168 hours. One indie dev told TechBloom his dating sim sat untouched for nine days before rejection arrived—because the AI had borrowed copyrighted dialogue it didn’t even know existed.
Apple pockets up to 30 % of every paid app, so more inventory should equal more cash. Yet the bottleneck is eroding goodwill at the worst moment. EU regulators already force sideloading; U.S. antitrust hawks circle. Delayed reviews send coders straight to TestFlight, enterprise certificates, or alternative stores on jailbroken devices—tiny cracks that widen into canyons if patience dies.
Friction turns into flight
Some studios now schedule “AI buffer weeks” the way airlines pad schedules for thunderstorms. Others ship stripped-down MVP shells, then push features after approval—gaming the very safety model Apple evangelizes. The result: users download placeholder apps that mutate overnight, precisely the sketchy behavior human review was meant to stop.
Apple won’t confirm headcount increases, but job boards list dozens of “App Review Analyst” slots in Austin and Cork. Contractors whisper of 12-hour shifts, eye-bleed quotas, and burnout so acute that internal error rates are climbing. Meanwhile, Google Play leans on automated scanners; Microsoft’s Store outsources bulk scanning to AI and spot-checks the edge cases. Apple clings to the moral high ground, yet the ground is sliding.
Expect a pivot, not a surrender. Cupertino is prototyping hybrid pipelines—machine pre-screen for plagiarism, private API calls, known malware signatures; humans keep the final veto. The code exists inside Apple already; the company uses it for internal betas. Rolling it out publicly would require admitting the perfect human filter is no longer tenable. Pride costs time, and time is now the enemy.
Bottom line: the same generative tools that let teenagers spit out Flappy Bird clones overnight are forcing the world’s most valuable company to choose between ideology and speed. If Apple clings to flesh, delays will stretch further, developers will defect, and the walled garden will start to look like a traffic jam. If it automates, it becomes just another app bazaar—safer than most, but no longer mythically different. Either way, the era of vibe coding has already decided the battle; Apple is only now counting the casualties.