Apple’s 2025 App Store report paints a picture of a heavily vetted marketplace. The company’s review team examined more than 9.1 million submissions, mixing human eyes with machine learning. Over two million apps and updates—about 1.2 million new titles and 800 thousand revisions—failed to meet the Review Guidelines and were rejected. Such rigorous filtering aims to reassure consumers wary after high‑profile phishing scams last year.

Apple says its Trust and Safety squads lean on AI to chase fraud tactics. In 2025 the firm terminated 193,000 developer accounts flagged for deception and blocked more than $2.2 billion in suspect transactions. Machine‑learning models flag anomalous code patterns and sudden permission spikes, giving reviewers a queue. The system sifted through 1.3 billion user reviews, excising roughly 195 million fraudulent or spammy entries before they reached shoppers.

These numbers illustrate the scale of moderation required to keep the storefront trustworthy. While fraudulent apps still slip through occasional cracks, the blend of AI triage and human expertise accelerates clean releases for legitimate developers. Apple’s data suggests the layered approach curtails abuse without throttling the volume of new software reaching iOS users.