At WWDC 2026 (keynote June 8), Apple replaced Siri with Siri AI: rebuilt with deep personal context, on-screen awareness, and — per reporting from TechCrunch and Tom’s Guide — Google Gemini under the hood. Siri AI ships as a standalone app in addition to working system-wide, can answer questions about what the iPhone camera sees, and rolls out with iOS 27 to devices as old as the iPhone 11.
What is confirmed vs what is not
| Claim | Status |
|---|---|
| Siri rebuilt as "Siri AI" with system-wide context + standalone app | Announced on the WWDC keynote, June 8, 2026 |
| Google Gemini powers Siri AI under the hood | Widely reported (TechCrunch, Tom’s Guide, CNBC); consistent with the Apple–Google deal reported earlier |
| Visual intelligence: ask Siri about what the camera sees | Announced |
| Eligibility from iPhone 11 onward, with iOS 27 | Announced |
| Exactly which queries route to Gemini vs on-device models, and whether web answers cite sources | NOT published by Apple — treat any specific routing claim as speculation |
Why this matters for brand discovery
- Distribution. Gemini-pipeline answers now reach the iPhone install base through the default assistant — buyers who never opened the Gemini app will get Gemini-flavored answers from their lock screen.
- Query shape changes. Voice and camera questions are longer and more conversational than typed ones: "what is this and is it any good" pointed at a product is a brand-visibility query with zero keywords. Put the exact product name plus its key spec in the first sentence of every product-page section (e.g. "the X200 runs 14 hours per charge") so a zero-keyword voice query can still be answered verbatim.
- One source of truth, two surfaces. If Gemini holds a wrong price or a stale claim about your brand, that error now surfaces both in the Gemini app AND through Siri — and correcting it means fixing the upstream source Google consumes (your product page or Merchant Center feed), not the assistant.
What brand teams should do now (4 steps)
- Baseline your Gemini answers this month, before the iOS 27 rollout: run your top 20 buyer prompts on Gemini and save the answers. When Siri AI lands, you will want a pre/post comparison.
- Audit accuracy on Gemini specifically: price, availability, return policy, key feature claims. Log every wrong claim with the prompt that produced it.
- Keep Google-side surfaces clean: Schema.org Product markup, Merchant Center feed accuracy, and consistent NAP/entity data — the structured signals Google systems already consume are the most plausible inputs to Gemini’s product answers.
- Add conversational phrasings to your tracked prompt set ("which <category> should I get if <constraint>", "is <brand> any good") — the voice-shaped queries Siri will forward.
Where Arenza fits
Arenza’s weekly scans cover Gemini today (alongside ChatGPT and Perplexity) on the Protect tier — $99/mo + usage, multi-brand, white-label — tracking visibility per prompt and flagging wrong claims with the producing prompt attached. Arenza re-runs the tracked prompt set on Gemini every 7 days and reports each changed answer with the prompt that produced it — so steps 1–2 and 4 become a standing weekly report. We do not measure Siri end-to-end (nobody can yet — Apple has not shipped it); we measure the Gemini layer the reporting says it stands on.
Sources
- TechCrunch WWDC 2026 recap: https://techcrunch.com/2026/06/09/wwdc-2026-everything-announced-on-siri-ai-os-27-apple-intelligence-and-more/
- Tom’s Guide, Apple Intelligence at WWDC 2026: https://www.tomsguide.com/ai/apple-intelligence-all-the-major-announcements-made-at-wwdc-2026
- CNBC live coverage: https://www.cnbc.com/2026/06/08/apple-wwdc-2026-live-updates.html
Facts checked 2026-06-11. The Gemini-under-the-hood claim is from press reporting, not an Apple spec sheet; routing details may differ at ship time. Email hello@arenza.ai with corrections and we will update within 48 hours.