Apple and Google say Apple Foundation Models based on Gemini will help power not only a more personalized Siri, but also “future Apple Intelligence features”
A Prediction That Came TrueWesley Hilliard /AppleInsider:No, Google Gemini will not be taking over your iPhone, Apple Intelligence, or SiriChris Velazco /Washington Post:Apple taps Google to power a new Siri and future AI toolsSamantha Subin /CNBC:Apple signs a multiyear Google deal to use Gemini models to power a more personalized Siri in 2026, saying Google's tech “provides the most capable foundation”John Gruber /Daring Fireball:Apple, Rather Quietly and With No Details, Announces Partnership
Context & Ripple Effects
Apple’s Gemini discussions were previously framed around a Siri overhaul; the subsequent multiyear Gemini agreement for a more personalized Siri made that supplier relationship concrete. This report broadens the stated scope from one assistant experience to Apple Intelligence features more generally.
The expanded framing matters because Apple is positioning Gemini-derived foundation models as an input to its own AI stack, rather than presenting Gemini as a separately branded consumer service. Reporting on the deal’s implementation also described Apple’s ability to fine-tune Gemini for its own use.
First-order effects
- Apple can apply Gemini-based Apple Foundation Models beyond the personalized Siri work, extending the deal’s relevance across future Apple Intelligence development.
- Google gains a larger role as a model supplier to Apple, while Apple retains the Apple Intelligence and Siri product layers rather than handing them to a Google-branded service.
Second-order effects
- The deal raises the strategic value of model access for platform companies: Apple can concentrate on product integration and customization while sourcing a capable foundation-model base from Google.
- Other AI model providers face a clearer distribution hurdle: Google’s models gain a route into Apple’s AI roadmap without Google needing to own the end-user interface.
Third-order effects
- If similar arrangements persist, consumer-device platforms may increasingly separate the customer-facing AI experience from the underlying frontier-model supplier, making distribution and integration as important as model ownership.
- That structure could concentrate leverage among a small set of model providers and major device platforms, while leaving branding, privacy posture, and feature control as key differentiators.
The trend: This is one data point in the shift toward major consumer platforms pairing proprietary AI experiences with externally supplied foundation models.