Siri is now an AI citation engine: what WWDC 2026 changed
Apple shipped Siri AI at WWDC on June 8, 2026, backed by a custom Gemini model. Here is what the iOS 27 launch means for brand citations across 1.8 billion devices and what to do before September.
By Stephan Charles | Last fact-checked: <time datetime="2026-06-12">June 12, 2026</time>
Apple shipped Siri AI at WWDC 2026 on <time datetime="2026-06-08">June 8</time>, backed by a custom 1.2-trillion-parameter Gemini model that Apple licenses from Google for roughly $1 billion a year. iOS 27 ships to 1.8 billion active Apple devices in September 2026. When it does, Siri becomes the ninth major AI platform brands need citations on. Through a new Extensions system, users can also route Siri queries to Claude or ChatGPT. BrandCited audits Gemini as one of its eight monitored engines, so brands that already appear in Gemini citations are positioned for Siri from day one. Brands that aren't cited in Gemini answers have until September to close that gap.
BrandCited's free scan checks your Gemini citation rate alongside seven other engines in 30 seconds. Run it at brandcited.ai.
Apple unveiled Siri AI on <time datetime="2026-06-08">June 8, 2026</time> at its annual Worldwide Developers Conference, replacing the previous version of Siri with an entirely new AI assistant powered by a custom Gemini model licensed from Google. The new Siri ships as a standalone app in iOS 27 alongside its existing voice interface, adds conversational memory across apps, and handles multi-step tasks including booking, searching, and summarizing without switching apps.
Two facts define the scope of this change. First, Apple licensed a custom 1.2-trillion-parameter model from Google at a reported cost of $1 billion per year, according to CNBC's WWDC live coverage and corroborated by TechCrunch's full WWDC recap. Second, through a feature called Extensions, users can set Claude, ChatGPT, or Gemini as their default AI engine, routing eligible Siri requests to that model instead of Apple's custom one. Developer betas opened the same day; public betas arrive in July and iOS 27 ships in September 2026.
Siri AI is not a minor update to an existing product. Apple rebuilt the core inference pipeline, replaced the previous NLU stack with a frontier model, and opened the assistant layer to three competing AI providers simultaneously. That combination creates a citation surface that didn't exist before June 8.
Why does Siri AI create a new citation surface for brands?#
Siri AI answers factual questions about products, companies, and services by generating prose responses — the same citation-dependent behavior that defines ChatGPT, Perplexity, and Gemini. Every iPhone user who asks Siri "what project management tool is best for a five-person team" or "which AI visibility platform tracks the most engines" gets an answer that either includes or excludes your brand name. That answer comes from Gemini's citation logic by default.
The scale matters. Apple has 1.8 billion active devices according to Apple's own WWDC announcement. iOS 27 adoption reaches 60% of eligible devices within 90 days of launch, based on Apple's historical update cadence. That translates to over a billion devices running Gemini-backed Siri by December 2026. No other single AI citation surface has reached that device count.
The Opollo 2026 AI Search Benchmark Report, which analyzed GA4 data from 312 B2B technology companies, found that AI-referred visitors convert at 14.2% versus 2.8% for traditional organic search. Siri AI will generate AI referral traffic at a scale organic search never approached on mobile. Brands cited in Siri answers don't just get impressions — they get high-intent traffic from users already in a decision frame.
A brand invisible to Gemini becomes invisible to Siri AI on the day iOS 27 ships. Siri defaults to the custom Gemini model, so Gemini's citation pool is Siri's starting citation pool. There's no secondary fallback for brands that Gemini doesn't mention.
The Extension system adds a compounding problem. Tom's Guide's analysis of Extensions reports that users can route Siri to Claude, ChatGPT, or Gemini on a per-query basis. Cross-engine citation overlap sits at approximately 11%, according to Authoritytech.io's citation analysis. A brand cited by Gemini has roughly a 1-in-9 chance of also appearing in Claude answers and a separate 1-in-9 chance of appearing in ChatGPT answers. Brands covering only one or two engines lose visibility on every user who routes Siri to a different model.
Even within Google's own products, citation pools diverge sharply. Authoritytech.io found that Google AI Mode and Google AI Overviews share only 13.7% of cited URLs. Siri's custom model, fine-tuned on Apple's data and conversational patterns, will develop distinct citation preferences over time, further separating its results from standard Gemini answers.
: If your brand scores below 40 on Gemini in BrandCited's audit, you have a measurable gap that iOS 27 will expose on 1.8 billion devices. The three months before September is the window to close it.
Run a free BrandCited scan to see your Gemini citation rate right now: brandcited.ai
Which content signals carry weight with the Gemini-backed Siri?#
Gemini's citation engine weights five signals that apply to Siri AI, based on Google's Knowledge Graph integration) and BrandCited's scan patterns across 2,000+ tracked brands.
Entity resolution is the first signal. Gemini cross-references brand names against Google's Knowledge Graph before generating an answer. A brand with a verified Knowledge Graph entity, consistent NAP data across the web, and a Wikidata entry gets surfaced at higher rates than an unresolved brand name.
Structured data extractability is the second signal. Gemini pulls passages from pages with Article, FAQPage, and Organization schema. A Moz 2026 analysis found that 88% of Google AI Mode citations come from pages outside the organic top 10, meaning schema-optimized content earns AI citations regardless of its search ranking.
Track your AI visibility for free
See how ChatGPT, Claude, Gemini, and 4 other AI platforms mention your brand.
Content freshness is the third signal. Gemini weights pages updated within the last six months over stale content, according to Digital Applied's Siri traffic analysis. A page not refreshed since 2024 carries lower citation probability regardless of its original quality.
Named author authority is the fourth signal. Anthropic's Claude and Gemini both weight byline prominence. Articles with a named author, a visible bio, and a LinkedIn sameAs in the JSON-LD Person schema earn more citations than anonymous content.
Corroborated claims are the fifth signal. Gemini weights claims appearing consistently across multiple authoritative sources. A brand described in industry coverage, analyst reports, and its own content earns higher citation confidence than a brand appearing only on its owned domain.
These steps are ranked by citation impact, highest first.
1Run a BrandCited scan to get your current Gemini citation score. You need a baseline before you can set a target. The scan is free and shows which queries you answer and which you miss across all 8 engines: brandcited.ai
2Check your Google Knowledge Graph entity. Search your brand name to see if Google has a knowledge panel. If it doesn't, create a Wikidata entry, ensure your Google Business Profile is complete, and add Organization schema with sameAs links to your official profiles.
3Add FAQPage schema to your top 10 pages. Gemini extracts FAQ content at the passage level. Write questions as the user would ask Siri: "What does [brand] do?" "How much does [brand] cost?" "Is [brand] good for [use case]?"
4Refresh your three most-cited pages. Update statistics, add a published/modified date in Article schema, and verify all external links resolve.
5Build coverage on Claude and ChatGPT. With Extensions, roughly one-third of Siri users will route queries to a non-Gemini model. BrandCited's scan shows your citation rate on all three engines so you can close the gaps that matter most for your category.
6Add a named author block with LinkedIn sameAs to every article. This single change lifts citation rates on Perplexity, Gemini, and Claude simultaneously.
BrandCited's audit engine checks for Gemini citation readiness as part of its Gemini engine score. The audit runs 30+ checks across entity schema, content structure, author authority, and citation frequency, and returns a score from 0 to 100 with each failing check ranked by impact. Since Siri AI defaults to Gemini, a brand's Gemini score in BrandCited is the closest available proxy for Siri readiness before iOS 27 ships.
Run a free BrandCited scan to see your citation rate across all 8 AI engines before iOS 27 ships. The scan takes 30 seconds and shows every gap ranked by impact: brandcited.ai
Google AI Overviews opt-out: Google's AI Overviews opt-out becomes enforceable on June 17, giving site owners six days to decide whether to stay in or exit AI-generated answers. (BrandCited coverage)
ChatGPT Lockdown Mode: OpenAI introduced Lockdown Mode for all logged-in ChatGPT users, limiting web access to reduce prompt injection risk. (OpenAI release notes)
Gemini 3.5 Pro timeline: Google confirmed Gemini 3.5 Pro is in final preparation for June release. (Search Engine Journal)
AI referral traffic growth: AI referral traffic grew 975% year-over-year as of Q1 2026, with AI-sourced sessions converting at 14.2% versus 2.8% for organic search. (Opollo 2026 AI Search Benchmark)
Perplexity Computer for Mac: Perplexity expanded its Personal Computer agent to Mac users this week, adding multi-model AI orchestration previously available only on Windows. (OpenTools)
Does Siri AI use the same citation pool as Google AI Mode?
No. Siri AI runs on a custom 1.2-trillion-parameter Gemini model Apple licensed from Google, not the standard Gemini used in Google Search. Authoritytech.io data shows Google AI Mode and Google AI Overviews share only 13.7% of cited URLs, indicating even Google's own AI products maintain separate citation pools. Siri's custom model will develop its own citation patterns as Apple fine-tunes it for conversational use cases.
Which AI model does Siri use by default in iOS 27?
Siri AI defaults to a custom Gemini model. Apple licensed it from Google at a reported cost of around $1 billion per year, according to CNBC and TechCrunch coverage of WWDC 2026 on June 8, 2026. Users can switch to Claude or ChatGPT through the Extensions system in Settings under Apple Intelligence and Siri.
When does iOS 27 ship to users?
Developer betas started on June 8, 2026. Public betas follow in July 2026. The public release is expected around September 14, 2026, based on Apple's historical fall release schedule.
How do I know if my brand is cited by Gemini?
Run a free BrandCited scan at brandcited.ai. The scan checks citation rate across 8 AI engines including Gemini and returns a score in 30 seconds, showing which queries your brand answers, which it misses, and a prioritized fix list.
Does optimizing for Gemini also optimize for Siri AI?
For the iOS 27 launch, yes. Siri AI defaults to a custom Gemini model, so content that earns Gemini citations carries into Siri answers. Over time, Apple will fine-tune its model and Siri will develop its own citation preferences. BrandCited tracks Gemini as one of its eight monitored engines and will add Siri as a separate surface when its citation behavior becomes distinguishable.
What happens to brands not cited by Claude or ChatGPT when users switch their Siri Extension?
Those brands become invisible on that user's device. Cross-engine citation overlap sits at approximately 11%, meaning a brand cited by Gemini has roughly a 1-in-9 chance of also appearing in Claude answers. The Extensions system turns this into a real gap: brands need consistent citation coverage across Gemini, Claude, and ChatGPT to cover all three engines users can route Siri to.
Track your AI visibility now
See exactly how AI engines like ChatGPT, Perplexity, and Gemini perceive your brand.
The BrandCited team covers GEO, AI search optimization, and brand visibility strategy. We publish research, practical guides, and product updates every week.
