Apple has quietly registered genai.apple.com, a generative-AI subdomain spotted on May 23 by MacRumors contributor Aaron Perris roughly two weeks before WWDC 2026 opens. The address is already configured on Apple's name servers but does not resolve — requests return a connection timeout rather than a 404, which signals the destination is provisioned but not yet switched on. Apple has publicly promised to detail "AI advancements" across its platforms at the conference, which runs June 8–12 with the keynote at 10 a.m. Pacific on June 8.
For practitioners, the subdomain itself is a footnote. The signal is that Apple is standing up a dedicated generative-AI surface separate from its existing Apple Intelligence page — and the developer-facing changes reportedly arriving alongside it.
A new framework layer for developers
Apple is reportedly preparing Core AI, a framework positioned as a successor to Core ML, with the two expected to run in tandem through a transition. Where Core ML shipped optimized on-device inference, Core AI is described as a higher-level interface for wiring generative models — including third-party LLMs — into apps. Pair that with the upgraded Foundation Models framework, the on-device model API Apple introduced at WWDC 2025 to give apps direct access to the same on-device model that powers Apple Intelligence, and the platform story becomes about giving developers a sanctioned path to embed generative features without standing up their own inference.
The other reported piece is AI Extensions: a mechanism for third-party chatbots and AI tools to plug into iOS so they can be invoked as default assistants and handle tasks Apple Intelligence would otherwise route internally.
The model layer: Gemini inside the stack
The capability gap is being filled by Google. Apple and Google reportedly struck a multi-year, non-exclusive partnership in January that places a custom Gemini model at the core of Apple's Foundation Models and powers the long-delayed Siri rebuild — which is expected to add on-screen awareness and multi-step, cross-app actions. Crucially, the arrangement is non-exclusive: users are expected to be able to select ChatGPT or Gemini as their preferred provider, and developers building on AI Extensions inherit that optionality rather than being locked to a single vendor.
What it changes for builders
If the rollout matches the leaks, Apple shifts from a closed "Apple Intelligence" feature set to a model-agnostic platform layer: an on-device foundation model for cheap, private inference; a framework (Core AI) to call it; and extension points for swapping in frontier models when on-device quality isn't enough. For anyone shipping on iOS or macOS, that reframes the build-vs-buy calculus around AI features — and turns Apple's enormous installed base into a distribution channel for whichever model wins the default-provider slot. The hard confirmation is two weeks out; until then, genai.apple.com is the clearest tell that WWDC 2026 is a platform release, not just a Siri demo.
