Back to News
Market Impact: 0.2

Report: iOS 27 to let Siri work with any third-party chatbot

AAPLGOOGLGOOG
Artificial IntelligenceTechnology & InnovationProduct LaunchesAntitrust & CompetitionConsumer Demand & Retail

Apple will enable Siri to route queries to third-party chatbots (ChatGPT, Claude, Google Gemini) in iOS 27, effectively ending its exclusive OpenAI tie. Users can select their preferred AI via an Extensions option in Siri settings and will be offered app links if a chatbot isn't installed; the changes are expected to be showcased at WWDC keynote on June 8. The move increases competition among AI providers and should modestly enhance iPhone AI functionality and user choice, but is unlikely to be a near-term earnings catalyst.

Analysis

Opening a major smartphone OS to multiple AI engines changes distribution economics more than model economics: the marginal cost to an AI provider to reach an extra iPhone user falls from expensive marketing to a settings-toggle UX decision, which can compress effective CPI by an order of magnitude and materially raise LTV for subscription-based models. Back-of-envelope: if a provider converts 5–10% of a 1B active device base to a $3–7/mo plan, that’s $1.8–8.4B incremental ARR before revenue share—enough to alter fundraising and pricing power for mid-tier models within 12–24 months. The biggest indirect lever is platform capture and placement economics. Apple can preserve or even monetize openness via curated placement, referral fees, or differential UX prioritization; a modest $0.50–$2/month per opting user fee would flip a large chunk of the ARR back to the OS owner while avoiding antitrust headlines. Conversely, incumbent AI-first players lose a soft default advantage that translates to ad and query volume erosion — a few percentage points of search/query share reallocation on iOS maps to multi-billion-dollar ad revenue swings for dominant search providers over 1–3 years. On the supply side, easier integration lifts short-term API/inference demand and therefore cloud GPU consumption, benefitting infra providers — but there’s path risk: if platforms push more on-device privacy processing, that capex tailwind could be blunted and latency/quality tradeoffs will gate mainstream adoption. Near-term catalysts are concentrated: product demos and SDK details will move sentiment within days; measurable user opt-in and active query share will be the material read-throughs over quarters. Contrarian read: the market’s knee-jerk framing as a pure loss for major cloud AI players underestimates two stabilizers — (1) incumbent model quality keeps voluntary opt-in high for a single provider, and (2) Apple’s ability to monetize discovery means the OS owner doesn’t have to choose between control and revenue. Net-net, the headline opens optionality rather than delivering an immediate winner-takes-all reallocation; the real winners will be those who convert distribution into sticky monetization and predictable margin expansion over 12–36 months.