- OpenAI is exploring legal action against Apple after the ChatGPT-Siri integration failed to deliver the subscriber growth it expected, according to Bloomberg.
- Apple required users to specifically say “ChatGPT” to invoke the feature and limited ChatGPT outputs to small windows, choices OpenAI views as deliberately restrictive.
- Apple has since signed a multibillion-dollar deal with Google for Gemini to power the new Siri, and iOS 27 will let users integrate other AI models including Anthropic’s Claude.
OpenAI took what an executive now calls a leap of faith on Apple. The landing broke both ankles.
Bloomberg reported this week that OpenAI has hired an outside law firm and is actively exploring legal options against Apple over a ChatGPT-Siri integration that has failed on nearly every metric the AI company cared about. The partnership, announced with fanfare at WWDC 2024, was supposed to embed ChatGPT across Apple’s ecosystem and generate what one OpenAI executive described as billions of dollars per year in new subscriptions. That has not come close to happening.
The frustration is specific. OpenAI believes Apple intentionally designed the integration to minimize ChatGPT’s visibility. Users must explicitly say “ChatGPT” when speaking or typing a command to Siri — a friction point that effectively buries the feature. ChatGPT responses appear in small, easily dismissible windows. There was no prominent placement, no default option, and according to SiliconANGLE, no money changed hands in the original deal. Apple simply takes a cut of whatever new ChatGPT subscriptions flow through its ecosystem.
The Leap of Faith That Wasn’t
An OpenAI executive, granted anonymity by Bloomberg to discuss the strained partnership, described how the deal came together: Apple framed it as comparable to the Google-Safari search integration — a default placement that sent massive traffic and revenue to its partner. OpenAI believed it. The executive said they were told to trust Apple without being given specifics about how the integration would actually work.
The result was something closer to a closet than a storefront. Siri summons ChatGPT only as a fallback for world-knowledge queries, and only after the user invokes it by name. The integration lives inside Image Playgrounds and Visual Intelligence as a background utility, not a front-and-center feature. OpenAI now says it has done everything from a product perspective while Apple has not made an honest effort.
Efforts to renegotiate have stalled. OpenAI declined to extend the partnership to cover Apple’s AI models — a decision that looks deliberate in hindsight. Apple is now paying Google billions for Gemini to power the revamped Siri, announced in January 2026, while iOS 27 opens the door to other AI models including Anthropic’s Claude. OpenAI got the brand association. Google got the money.
Apple’s Partner Problem Has a History
This is not a new pattern for Apple. The company has a documented history of entering partnerships to solve short-term product gaps, then sidelining those partners once internal capabilities catch up or better terms emerge elsewhere.
Google learned this with Apple Maps. After relying on Google’s mapping data for the original iPhone, Apple launched its own maps app in 2012 and pushed Google’s offering off the default home screen. Qualcomm lived through years of patent litigation before Apple settled and bought Intel’s modem division. Epic Games discovered that Apple’s App Store terms were non-negotiable the hard way, resulting in a lawsuit that reshaped app store economics but left Apple largely in control.
OpenAI’s situation is distinct in one way: the partnership was never supposed to be exclusive, which means Apple’s deal with Google isn’t itself a breach. The breach, if one exists, would center on whether Apple met the promotional and integration commitments that OpenAI says were implied — even if they weren’t spelled out in the contract.
There is tension on Apple’s side too. OpenAI’s recruitment of former Apple engineers, including those working with Jony Ive on a hardware device, has reportedly left Apple executives fuming. Apple sees a partner that took the integration deal, used it for brand credibility, and then started building a competing product while poaching talent.
OpenAI will likely delay any formal legal action until after its ongoing court battle with Elon Musk concludes. A verdict in Musk’s nonprofit-conversion trial — which alleges OpenAI violated charitable-trust law when it restructured as a for-profit company — could come as soon as next week, with the jury set to begin deliberations Monday. Separately, Musk’s xAI has filed a different antitrust lawsuit alleging the Apple-OpenAI deal violates competition law, but the growing rift between the two companies may undermine his argument that they are colluding to lock out competitors.
The entire episode illustrates a structural problem for AI companies that depend on platform distribution: the platform holder always holds the better cards. OpenAI built the product, but Apple controls the surface where users encounter it. As one source close to the situation put it, the deal sounded amazing — until OpenAI realized it had brought the food to a restaurant where someone else owns the menu.
FAQ
What is OpenAI’s legal argument against Apple?
OpenAI believes Apple failed to honor the spirit of their partnership by designing the ChatGPT-Siri integration to be difficult to find and use. A potential breach-of-contract claim would center on whether Apple’s implementation met the promotional commitments OpenAI says were implied, even if not explicitly detailed in the contract.
Did money change hands in the Apple-OpenAI deal?
No. Per SiliconANGLE’s report, no payment was made as part of the original partnership. Apple takes a commission on ChatGPT subscriptions generated through its ecosystem, but the deal itself was distribution-for-brand, not a paid placement.
Why is Apple switching to Google Gemini for Siri?
Apple signed a multibillion-dollar deal with Google for Gemini to power the new Siri after OpenAI declined to extend the partnership to cover Apple’s AI model development. The move gives Apple a more willing AI partner and deeper integration capabilities.
What happens to Musk’s lawsuit against OpenAI and Apple?
In a separate antitrust lawsuit filed by Musk’s xAI, Musk alleged the Apple-OpenAI deal violated antitrust law by conspiring to dominate the chatbot and smartphone markets. That lawsuit survived motions to dismiss, but the increasingly public rift between OpenAI and Apple may undermine Musk’s claim that the two companies are colluding.
[Editor’s note: This article was updated on May 16, 2026 to correct two errors. (1) The Apple-Google Gemini deal was announced in January 2026, not at WWDC 2026, which had not yet occurred. (2) The article conflated two separate Musk lawsuits — the nonprofit-conversion trial (Musk v. Altman, with a verdict expected soon) and the xAI antitrust lawsuit (which specifically alleges the Apple-OpenAI deal violated competition law). The text has been revised to clearly distinguish between these two cases. Additionally, an incorrect internal link regarding Jony Ive was removed.]
