Apple signed a partnership with Google to use Gemini for a revamped Siri, but a new report says Apple once leaned in a different direction. Bloomberg reporter Mark Gurman says Apple explored rebuilding Siri around Anthropic’s Claude, and it already relies on Anthropic models for some internal work.
Gurman, in an interview on TBPN, said Apple had serious talks about using Anthropic’s Claude as the core model for a rebuilt Siri. He also said Apple uses Anthropic in internal product work and tools, including custom deployments, before Apple finalized the Gemini agreement.
The key point: Apple did not treat Gemini as its only option. It tested multiple outside models and weighed tradeoffs before it landed on Google.
Why Apple did not pick Claude for Siri
According to a separate report by The Information, Anthropic sought “several billion dollars” over multiple years to supply models for a Siri rebuild. The report said the pricing became a sticking point.
That helps explain why Apple kept looking, even while it continued working with Anthropic in other areas.
Where Google Gemini fits now
Apple and Google announced a multi-year partnership to integrate Google’s Gemini AI models into a revamped Siri in 2026. Apple still positions privacy as central, leaning on on-device processing and Private Cloud Compute for certain workloads.
So if you use Siri later in 2026, you may end up using a system that blends Apple software, Apple privacy controls, and Google’s model tech behind the scenes.
If Gurman’s reporting holds, Apple’s Siri roadmap looks less like a single straight line and more like a series of pivots.
- Apple wants a big Siri jump, not a small patch.
- Apple treats top AI models as suppliers, and it will switch when the terms or capabilities do not match Apple’s needs.
- Apple still works with multiple AI partners, even if one partner becomes the headline name for Siri.
