If it feels harder to ship an app right now, it's because it is. App Store releases in April 2026 are up 104% year-over-year across both stores — 89% on iOS alone. And review teams are visibly tightening the net.
what's actually happening#
TechCrunch reported last week that the submission flood is downstream of AI coding tools becoming good enough for non-developers to ship their first apps. Claude Code, Replit, Cursor, and the whole cohort of agentic dev tools have crossed a usability threshold where someone with an idea and zero Swift background can produce a submittable binary in a weekend.
Great for the ecosystem. Less great for your odds of getting through review cleanly. Reviewers are seeing volumes they were never staffed for, and they're triaging by rejecting the easy cases aggressively.
the three buckets getting rejected#
Dynamic code execution. Any app whose core behavior is shaped by code fetched or generated after install is getting flagged. Includes plenty of AI apps where the "feature" is "prompt an LLM and render the response" — if the response can change how the app behaves, Apple reads that as dynamic code.
AI disclosure gaps. If your app sends user data to an external AI provider — OpenAI, Anthropic, Google, anyone — you need a consent screen that names the provider and explains what's being shared. "We use AI to improve your experience" is not enough. Reviewers are checking.
Template-wrapper thinness. Apps that are visibly a thin shell around a web view, a prompt box, or a single AI call without real product scaffolding are being rejected under 4.2 Minimum Functionality. The bar moved. A novelty wrapper that would have slipped through in 2024 reads as slop in 2026.
what this means if you're not a vibe-coder#
The awkward part: legitimate indie devs are caught in the same slowdown. Reviewers spending more time rooting out low-effort AI submissions means your 14-day resubmission cycle just became a 21-day cycle.
The signal your listing sends matters more than ever. Generic screenshots, descriptions that read like ChatGPT one-shotted them, icons that look like DALL-E presets — these now correlate in the reviewer's head with the low-quality submissions they're trained to reject.
what to do about it#
Ship fewer, tighter submissions. Every rejection resets your queue position — one clean submission beats three iterative ones. Over-disclose your AI usage anywhere it appears. And make your listing look like a human made it, because the AI-generated ones in reviewers' queues right now are mostly bad.
where stora fits#
Stora's compliance engine runs your binary and metadata through the exact rejection-pattern checks reviewers use — dynamic code flags, AI disclosure gaps, privacy manifest completeness, minimum functionality heuristics. If any of those would trip review, you see it before you submit, not after. The build repair agent patches the common ones automatically.
This used to be a nice-to-have. At 104% submission growth and tightening review standards, it's the difference between shipping this sprint and shipping next month.
The App Store is busier than it's ever been. That's good for users and bad for carelessness. Tighten the submission. Ship the thing.