Agentic AI Lifts iOS App Releases 30% as Usage Stays Flat
Updated
Updated · InfoWorld · Jun 8
Agentic AI Lifts iOS App Releases 30% as Usage Stays Flat
1 articles · Updated · InfoWorld · Jun 8
Summary
More iOS apps are reaching market, but user traction is not following: a new NBER working paper found releases surged after agentic AI spread while reviews fell and significant-usage apps stayed essentially flat.
100,000-plus GitHub developers tracked in the study showed AI sharply boosted coding output—commits rose 40% with autocomplete, 140% with interactive agents and 180% with autonomous agents—but that translated into only about 50% more projects and 30% more releases.
Researchers call it a weak-link problem: AI strengthens code writing far more than product design, integration, maintenance and trust-building, with an elasticity of substitution of 0.25 indicating AI complements rather than replaces human work.
Across four major app marketplaces, the paper found more new apps but no increase in total usage, suggesting attention, distribution and go-to-market execution—not code generation—are becoming the main bottlenecks.