If everyone’s doing vibe-coding

Apple’s App Store has an ever-increasing number of apps, and it’s taking longer and longer to carry out the necessary reviews before publishing them: this is down to AI, of course, which allows anyone to create (or have someone create?) an app, and then offer it to the general public. But… does anyone actually download these apps in the end?

1 NOV 26
Translated by AI
Image of If everyone’s doing vibe-coding

Photo via Getty Images

On its website, Apple states that the review process for apps on its App Store is swift: “typically, 50 per cent of submissions are reviewed within 24 hours; 90 per cent within 48 hours”. This has long been the hallmark of the App Store: an application is developed and submitted to Apple, which, within a couple of days, reviews it, verifies that everything is correct and secure, and makes it available worldwide. In return, the company charges a 30 per cent commission on any transaction made through an app downloaded in this manner (i.e., any app for iPhone, iPad, and so on). This business model is highly profitable for the company, but it is deeply resented by developers, who have been criticising the ‘Apple Tax’ for years and have already caused the company a number of legal problems.
This controversial agreement between Apple and developers has been put to the test by the rise of artificial intelligence (AI) systems capable of generating code. It is now possible to open Claude Code and ask it to develop an app for taking notes or drawing, for example, and let the machine do the rest. Anyone can do it: when people who do not know how to programme (or who, in any case, do not intend to review the generated code in any great detail) interact with AI in this way, the term ‘vibe coding’ is used.
The coding craze has transformed many sectors, including the App Store. According to data from Sensor Tower, a service that monitors app traffic, in 2025 the number of new apps published on Apple’s App Store grew by 30 per cent, reaching around 600,000. This figure has practically doubled over the course of 2026: in the first half of the year, 560,000 were published, just under the total released throughout 2025, as reported by the New York Times.
The picture is clear: if anyone can create an app, more and more people will give it a go – for fun, for study, out of curiosity, or in the hope of landing a hit. In most cases, these are ‘clone’ services that perform poorly and, above all, nobody downloads: last year, app downloads rose by just 3 per cent; in the first half of 2026, by 2 per cent. In short, apps created using vibe coding do exist; they are amongst us, but only in potential; in reality, few people use them, and they merely add to a catalogue that is increasingly full of forgettable offerings.
Despite this, these apps are uploaded to the App Store, which subjects them to the necessary review. And this brings us to the issue of timing. The Register website has in fact compiled a number of posts from an Apple developers’ forum, in which users complain about how the process has become lengthy and tortuous. Forget a 48-hour wait to see your app go live: there are people who have had to wait two or three weeks; one person claims to have been waiting for over a month.
This is the consequence of ‘slop’, the term used to describe images, videos and sounds generated by AI and considered to be of poor quality. Rather than being synonymous with ugliness, however, ‘slop’ is about speed and immediacy: from a certain point of view, a lack of effort and direction, because the AI does everything; the user simply has to write the ‘prompt’. Meanwhile, however, the infrastructure required to distribute the software remains that of the pre-ChatGPT ‘ancien régime’. The same is happening to the World Wide Web and, on closer inspection, to society as a whole.
The risk for the App Store is that it could become a shop where there is suddenly a queue at the entrance, but everyone is simply ‘window-shopping’, as the saying goes, and revenue does not rise as one might expect. Not that this is an existential problem for Apple, which is worth several thousand billion dollars and, above all, as mentioned, has been monetising the App Store rather aggressively for decades.
Of course, we imagine that Apple will use (if it doesn’t already) AI itself to speed up the review process, but apparently that isn’t enough. It is the agreement between developer and company that has been broken, now that the term ‘developer’ potentially includes all of us and the App Store has become a place where anyone can submit their app, just like any other. This is not exactly the ideal formula for a service that is supposed to guarantee security and reliability – and which, on top of that, requires the payment of a substantial fee.