Apple’s Siri Will Use Every Chatbot for Smarter AI Answers

Apple has already connected Siri to ChatGPT, but the company’s broader AI strategy now points to a routing model in which Siri can draw on multiple outside systems rather than rely on a single chatbot. Apple formally announced ChatGPT integration on June 10, 2024, and support documents now show that ChatGPT works inside Siri and Apple Intelligence features. Separate reports from Bloomberg, echoed by MacRumors through 2025 and 2026, indicate Apple has also evaluated Anthropic and Google Gemini as it rebuilds Siri’s delayed large-language-model stack.

That matters because Apple is no longer framing Siri as a standalone assistant that must answer everything with one in-house model. Instead, the company appears to be building Siri as an orchestration layer: on-device models for privacy-sensitive tasks, Apple-run models for broader reasoning, and third-party chatbots when outside expertise is useful. Apple’s own product pages say Siri can use ChatGPT “when helpful,” while preserving user consent before requests are sent. Reports over the past nine months suggest that this architecture may expand well beyond OpenAI.

June 2024 to March 2026: How Siri shifted from one partner to many

Apple’s public starting point is clear. On June 10, 2024, Apple and OpenAI announced that Siri would be able to hand off certain requests to ChatGPT powered by GPT-4o. OpenAI said users would be asked before questions, documents, or photos were sent, and Apple’s current support flow still includes a setting to confirm ChatGPT requests or disable the integration entirely. Apple’s product page continues to present ChatGPT as an optional extension inside Apple Intelligence rather than Siri’s only brain.

The more important shift came after Apple’s own Siri upgrade slipped. Bloomberg reported on June 30, 2025 that Apple was considering Anthropic Claude or OpenAI models to power a new Siri, potentially sidelining more of its in-house approach. MacRumors later reported, citing Bloomberg-based developments and subsequent commentary, that Apple had explored Claude, then moved toward a Gemini-linked path for parts of the next-generation Siri effort. By February 2026, MacRumors described a future in which Apple’s AI features would use a custom model built in collaboration with Google’s Gemini team, while still positioning Siri as a step toward chatbot-style interaction.

What is driving Apple’s multi-chatbot Siri design?

The answer is partly technical and partly strategic. Apple wants Siri to handle personal context, on-screen awareness, and app actions, but those features have been delayed into spring 2026, according to Bloomberg reporting summarized by MacRumors. That delay suggests Apple’s internal models alone have not met the reliability bar needed for a mass-market assistant embedded across iPhone, iPad, and Mac.

A multi-model design gives Apple several advantages. First, it lets the company keep simple or private tasks on device. Apple’s Apple Intelligence page says any app can tap into on-device models and that those features can work offline. Second, it allows Apple to use external frontier models for broader reasoning or knowledge-heavy prompts. Third, it reduces dependence on one supplier in a market where OpenAI, Anthropic, and Google are improving at different speeds.

For users, the practical implication is straightforward: Siri may become less like a single assistant with one personality and more like a broker that chooses the best engine for the task. Apple has not publicly confirmed a full “all chatbots” framework in those words. But the combination of confirmed ChatGPT support and repeated reports about Anthropic and Gemini makes that direction increasingly hard to ignore.

ChatGPT vs. Claude vs. Gemini: Why Apple may not pick just one

OpenAI has the only officially confirmed Siri integration today. Apple’s website and OpenAI’s help documentation both show that ChatGPT can be enabled in Siri settings, used with or without an account, and invoked directly by users. That makes OpenAI the incumbent external model inside Apple Intelligence.

Anthropic enters the story through reporting, not product pages. Bloomberg reported in mid-2025 that Apple was considering Anthropic Claude or OpenAI to power a more capable Siri. MacRumors later said Bloomberg’s Mark Gurman disclosed that Apple had at one point been close to rebuilding Siri around Claude before the strategy changed. That matters because Claude is widely used in coding and enterprise workflows, areas where Apple has been under pressure to improve its AI credibility.

Google Gemini appears to have gained momentum later. MacRumors reported in August 2025 that Apple was in talks with Google to use Gemini for the upcoming smarter Siri, and by March 2026 its Gemini guide said Apple CEO Tim Cook had reiterated the structure of a partnership with Google for the personalized Siri effort. Apple has not published a consumer-facing page confirming Gemini inside Siri in the same way it has for ChatGPT, so that part remains report-based rather than fully productized.

Spring 2026 is the key date for Apple’s delayed Siri overhaul

The next major checkpoint is spring 2026. MacRumors, citing Bloomberg, reported in June 2025 that Apple planned to release delayed personalized Siri features in spring 2026. Those features include personal context, on-screen awareness, and deeper app integration. If Apple meets that window, it will be the first real test of whether Siri can combine Apple’s privacy-first architecture with outside chatbot intelligence at scale.

That launch matters beyond Apple’s ecosystem. Amazon, Google, Microsoft, and Samsung are all pushing assistants that blend proprietary systems with large language models. Apple’s version appears more modular. Instead of betting everything on one chatbot, it is building a layer that can switch among models, ask permission, and keep some processing local. In competitive terms, that may be slower to ship, but it is also more flexible.

Frequently Asked Questions