OpenAI’s Siri alliance frays over money and control. Summary OpenAI is reportedly considering legal action against Apple over how ChatGPT has been integrated and monetized inside Siri on iPhones and other Apple devices. People familiar with the deal say OpenAI is disappointed that the integration has not produced the prominence or paid subscription conversions it expected, and that Apple does not pay OpenAI a direct licensing fee beyond a share of in‑app subscription revenue. OpenAI’s legal team has brought in an outside law firm to evaluate options, including sending Apple a formal breach‑of‑contract notice, while Apple is said to be sounding out rival AI providers for future Siri integrations. OpenAI is exploring legal options against Apple over their two‑year‑old deal to weave ChatGPT into Siri and iOS, amid frustration that the integration has failed to deliver the subscriber growth and commercial upside the startup expected. According to Bloomberg, OpenAI’s lawyers are “actively working with an outside legal firm on a range of options that could be formally executed in the near future,” including a breach‑of‑contract notice that would stop short of immediately filing suit.[3][5][7][1][2] The partnership, announced at Apple’s WWDC in 2024, allowed Siri to hand off complex queries to ChatGPT and folded OpenAI’s model into iOS “Visual Intelligence” features, where users can point their camera at objects or documents and ask questions that are routed to ChatGPT. At the time, reporting from Bloomberg and Fortune indicated that Apple “isn’t paying OpenAI” a direct fee for the integration, with both sides instead betting on converting a slice of Apple’s user base into paid ChatGPT subscribers at $20 per month and splitting in‑app revenue via the usual App Store rules. Less than two years later, that bet appears to have misfired. Sources quoted by Bloomberg say OpenAI believes it “fulfilled its obligations under the partnership, while Apple failed to deliver on key commitments,” particularly around how prominently ChatGPT is presented in Siri and how easy it is for users to discover and upgrade to paid plans. A report from Firstpost adds that OpenAI had expected Apple’s ecosystem to “drive billions of dollars in subscriptions” but has instead seen “lower‑than‑expected revenue and limited visibility” for ChatGPT features buried inside iOS settings. Traffic, ownership and the platform‑AI power struggle The core of the brewing dispute is not the quality of Siri’s answers, but who owns the traffic and economics once an AI model is embedded inside a dominant platform. Earlier reporting from Bloomberg made clear that Apple’s strategy was to offer ChatGPT “for free” as an optional Siri upgrade and then seek “revenue‑sharing agreements” where it takes a cut of any subscriptions sold on its devices, rather than pay model providers upfront. OpenAI is now reportedly asking whether Apple’s minimal promotion of ChatGPT within Siri and Settings — “functionality is embedded but does not lead to subscription conversion,” in the words of one legal summary — amounts to a breach of commercial expectations written into the deal. Apple, for its part, is said to have its own concerns. TechCrunch and Bloomberg both report that Apple executives are uneasy about OpenAI’s privacy posture and its push into hardware — particularly projects involving former Apple design chief Jony Ive — and are actively exploring deeper integrations with rival models from Google’s Gemini and Anthropic. For now, neither Apple nor OpenAI has publicly confirmed any legal move, and people close to the talks say OpenAI may wait until after its high‑profile trial with Elon Musk concludes before escalating against Apple. Behind the legal positioning is a bigger structural question: in a world where AI capabilities are embedded deep inside operating systems and voice assistants, does the value accrue to the model provider or the platform that owns the user, the interface and the billing relationship? OpenAI’s frustration with Apple’s “commercialization” of ChatGPT inside Siri is the first high‑stakes test of that question — and whatever answer emerges will shape how every future AI–platform alliance is negotiated.
