Apple Overhauls Intelligence with Google‑Based AI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Apple announced a major overhaul of its Apple Intelligence platform, introducing a new architecture built on foundation models developed with Google that can run on-device and on cloud servers.
- Apple Intelligence is being rolled out across everyday iOS experiences, as highlighted in Apple’s newsroom release.
- Apple expanded its Private Cloud Compute beyond its own data centers, enabling broader AI workloads.
- Apple is using Google’s Gemini models as a basis for its new AI stack, but not the same version deployed by Google, per a Mark Gurman tweet.
- Siri’s advanced AI model will be available only on the iPhone 17 Pro and will not be offered in the EU, according to a user tweet.
- Apple Intelligence can generate iOS shortcuts from natural‑language descriptions, as demonstrated in a user tweet.
- Apple’s new AI models employ a mixture‑of‑experts architecture with a 3‑billion‑parameter dense core and a 20‑billion‑parameter total, using early routing to activate experts per prompt.
Why it matters: Apple’s new AI stack gives it a proprietary, privacy‑centric advantage that can drive premium device sales and lock users into its ecosystem, while EU consumers and voice‑dictation rivals miss out on the most advanced Siri capabilities, and the broader AI market sees a new competitor with cloud‑edge integration.
