The report, which analyzes over $11 billion in annual revenue across 75,000 developers, indicates that AI integration is now a mainstream phenomenon, accounting for 27.1% of all subscription apps. The Photo & Video category leads the charge, with 61.4% of its apps now marketing AI features. This surge in adoption is fueled by a superior ability to monetize downloads; AI apps generate 2.4% conversion from download to paid status, compared to 2.0% for non-AI apps. However, this early success is being undermined by a retention crisis. Annual retention for AI apps sits at a meager 21.1%, nearly ten percentage points lower than the 30.7% seen in the broader market.
The volatility of the AI sector is further evidenced by a 20% higher refund rate for AI-powered products. RevenueCat notes that median refund rates for AI apps hit 4.2%, with the upper bound reaching a staggering 15.6%. This suggests a "buyer's remorse" cycle where users, enticed by aggressive marketing or the promise of a magic-bullet solution, find the actual output lacking or redundant after a few uses. The rapid pace of technological advancement also plays a role, as users frequently hop between competing platforms to chase the latest model updates or feature sets, treating AI tools as disposable utilities rather than long-term partners.
U.S. President Trump has frequently emphasized the need for American dominance in the AI race, and the commercial data reflects a gold-rush mentality among developers. Yet the "app middle class" is thinning. While the top 10% of subscription apps saw their monthly recurring revenue grow by 306% over the past year, the share of new apps crossing the $1,000 monthly revenue threshold actually slipped from 19% in 2024 to 17% in 2025. For AI developers, the challenge is no longer proving that people will pay for the technology, but proving they will stay for it.
The only bright spot in retention for AI apps is at the weekly subscription level, where they outperform non-AI apps with a 2.5% retention rate versus 1.7%. This suggests that AI is currently viewed as a "project-based" tool—something a user pays for to complete a specific task, like editing a batch of photos or drafting a report, before letting the subscription lapse. To bridge the gap to annual sustainability, developers are beginning to pivot away from generic wrappers toward deeply integrated vertical solutions that solve persistent, recurring problems. Without this shift, the AI app boom risks becoming a high-conversion, high-churn treadmill that burns through venture capital and user trust with equal speed.
Explore more exclusive insights at nextfin.ai.
