- New app submissions to Apple’s App Store grew approximately 30% in 2025 to nearly 600,000, reversing a decline that had seen submissions fall 46% between 2016 and 2024, according to Sensor Tower data cited by The Information.
- Anthropic’s Claude Code and OpenAI’s Codex were identified as primary contributors, enabling non-programmers to build apps via written prompts.
- Apple’s App Review team is now processing more than 200,000 submissions per week and has deployed AI tools internally to help scale the review process.
- Apple pulled or blocked updates to vibe-coding platforms including Anything and Replit for generating interpreted code that alters an app’s primary purpose post-submission, violating App Review Guidelines.
What Happened
New app submissions to Apple’s App Store grew approximately 30% in 2025 to nearly 600,000 — reversing a decade-long contraction — according to a report published by 9to5Mac on April 6, 2026, citing an analysis by The Information based on Sensor Tower data. The rebound follows a period in which new app submissions fell 46% between 2016 and 2024. The Information attributed the reversal primarily to AI-assisted development platforms — referred to as “vibe coding” tools — including Anthropic’s Claude Code and OpenAI’s Codex.
Why It Matters
The multi-year decline in new App Store submissions pointed to rising barriers for independent developers trying to build and ship iOS software. The 2025 increase suggests that AI coding tools have materially altered those conditions. The shift is notable not just for volume but for the composition of who is now building: tools like Claude Code and Codex have, according to the report, made app development accessible to people without formal programming backgrounds, while also allowing experienced engineers to produce code at substantially higher throughput.
Technical Details
Abraham Yousef, senior insights analyst at Sensor Tower, directly linked the submission surge to agentic development platforms. “We’ve seen an explosive growth of new apps over the past year,” Yousef said. “It aligns with a broader release of agentic coding tools that remove prior difficulties of creating apps.”
The Information acknowledged that “it’s difficult to determine how many of these new apps developers created using AI,” but assessed that “it’s likely most rely on AI tools given how quickly they have been adopted.” Apple’s App Review team is currently processing more than 200,000 submissions per week — measured across the 12 weeks prior to the report — with 90% of submissions cleared within 48 hours and an average review time of 1.5 days. Apple confirmed to The Information that it is using AI tools internally to assist reviewers with the increased volume.
Who’s Affected
Non-programmers who previously lacked the resources to commission app development represent a new participant class in the App Store ecosystem. However, Apple has moved to restrict a subset of AI-powered development platforms — specifically those generating interpreted code that enables an app to change its own primary purpose after submission, which violates App Review Guidelines and the Developer Program License. Apps from Anything and Replit were among those pulled or blocked from receiving updates as a result of these enforcement actions.
Elon Musk posted on X in recent weeks that “iOS App Review delays are getting ridiculous.” Apple disputed the characterization, citing the 1.5-day average review time and the more than 200,000 weekly submissions now being processed.
What’s Next
Apple updated Xcode in recent months to support coding models and agents, though that integration is oriented toward experienced developers rather than the no-code segment now driving much of the new submission volume. Whether Apple revises its App Store policies ahead of WWDC26 to formally address AI-generated interpreted code — the technical category at the center of the Replit and Anything enforcement actions — has not been announced.
