Apple opens its AI to developers, while OpenAI makes 10 billion with ChatGPT
The objective is clear: to democratise access to on-device artificial intelligence, while keeping intact the pillars of the Cupertino company's philosophy, namely privacy and free service
2' min read
2' min read
During the ongoing WWDC 2025 in Cupertino, Apple announced a long series of novelties concerning its software ecosystem (from iOS 26 to the new Liquid Glass design). But above all, it announced an important new feature regarding its artificial intelligence: the Foundation Models Framework, a new API designed to allow third-party developers to integrate the language models underlying Apple Intelligence directly into their apps. What does this mean? It means that Apple's AI will be able to run, potentially, in any app installed on an iPhone, iPad or Mac. Developers and users will be able to take advantage of powerful AI capabilities offline, without depending on the cloud or external servers.
The objective is clear: to democratise access to on-device artificial intelligence, while keeping intact the pillars of the Cupertino company's philosophy, namely privacy and free service.
"We are giving developers direct access to the core on-device model that powers Apple Intelligence," said Craig Federighi, Apple's senior vice president of software at the event. "We think this can trigger a new wave of intelligent experiences within the apps that users use every day."
A few examples; with this framework, an educational app could generate customised quizzes from the student's notes, directly on the device. Or a hiking app could offer a natural trail search even without an internet connection.
It will be interesting to see how developers will interface with this novelty, and above all what Apple's business plans might be, which on the stock exchange - during the event - remained rather flat, registering -0.7%.


