Rewind the clock two years, and Apple had just officially integrated ChatGPT into the iPhone at WWDC (Apple Worldwide Developers Conference).
The scene at the time seemed like a perfect match: Apple possessed the world's largest and most expansive device empire, while OpenAI was responsible for catching up with Siri, which had been slow to learn to speak human language. Apple lacked models, and OpenAI lacked entry points; both sides got what they needed, a perfect match.
But today, Apple has sued OpenAI.
But today, Apple has turned around and taken OpenAI to court.
On July 10, 2026, Apple filed a lawsuit in the U.S. District Court for the Northern District of California against OpenAI, OpenAI's hardware company io Products, and two former Apple employees, Chang Liu and Tang Yew Tan. The lawsuit accused them of systematically stealing Apple's confidential documents, hardware components, manufacturing processes, and unreleased product information through recruitment, interviews, departure guidance, and supply chain collaborations, in order to accelerate OpenAI's consumer hardware business. Business warfare is often quite straightforward. So-called strategic partnerships are often just a polite way of putting things before the businesses of the two companies directly clash. Once the competition begins for the same users, talent, supply chains, and next-generation hardware entry points, even the closest allies can turn against each other. Moreover, OpenAI's expansion strategy over the past few years has almost always involved stepping into the heart of a tech giant's territory with each step forward. However, upon closer inspection, it seems OpenAI has a grudge against everyone. This company possesses an almost magical ability—it always manages to first shake hands and cooperate with others, then expand its business to their doorstep, ultimately turning partners into competitors and investors into its loyal "family," constantly on guard against OpenAI. This article provides a brief overview of OpenAI's love-hate relationships with leading tech companies. Apple: From Sweetness to Legal Battles
Apple: Once Sweet, Now in Court
The relationship between Apple and OpenAI is a classic example of a whirlwind marriage and divorce in the AI world.
In 2024, Apple was clearly a step behind in generative AI. ChatGPT had already firmly captured the minds of users worldwide, Google was pushing Gemini, and Microsoft was cramming Copilot into Office to promote office assistance. Meanwhile, Apple, the former "AI elder statesman" of the Chatgot era, was still struggling to prove on stage at its developer conference that Siri could indeed understand human speech. There was no other way but to seek external help. So Apple integrated ChatGPT into Apple Intelligence. When Siri encounters a question it can't answer, it can politely ask, "How about I ask ChatGPT for you?" OpenAI thus gained access to one of the world's most valuable consumer electronics gateways; Apple, on the other hand, temporarily masked its own shortcomings in large-scale model capabilities. However, in reality, the two companies never wanted the same thing from the beginning. Apple hoped ChatGPT would be a plugin, ideally quietly residing deep within the system, invoked when needed and kept out of the way when not. Apple controls users, hardware, the system, and distribution, while OpenAI provides some capabilities. But OpenAI's ambitions are clearly more than just being a high-end outsourcing solution residing within the iPhone; it aspires to become a new entry point for users. If in the future users no longer need to open apps, but instead directly tell AI, "Do this for me," then whoever controls the AI assistant may be able to bypass the traditional operating system. OpenAI later discovered that Apple was truly appealing, bringing in Jony Ive's team—one of the key figures behind iPhones, iPads, and MacBooks—to aggressively enter the consumer hardware market. Naturally, Apple couldn't sit still. They remained silent about the past relationship between Siri and Chatgpt, and now, while incorporating Gemini into the core capabilities of the new Siri, they've also sued OpenAI. According to Apple's lawsuit, the matter is not as simple as "normal poaching." Apple claims that a former employee, Chang Liu, after leaving to join OpenAI, still accessed Apple's internal servers through vulnerabilities and downloaded a large amount of confidential engineering data. He is also accused of helping other Apple employees copy files, evade security reviews, and prepare for OpenAI interviews. Apple further claims that OpenAI's recruitment system instructs Apple employees on how to handle departure reviews, reminding them not to reveal their future plans too early (especially not to mention going to OpenAI), not to sign documents casually, and to extend system access as much as possible. Even during interviews, former Apple employees at OpenAI use internal Apple project codenames, requiring candidates to prepare "in-depth technical analyses" of their current work. Some candidates are even asked to bring Apple batteries, circuit boards, logic boards, and prototype components to the interview. Of course, as of now, these are still Apple's unilateral accusations, and their ultimate validity depends on discovery of evidence and court proceedings. But regardless of the outcome of the lawsuit, the relationship between Apple and OpenAI is unlikely to return to what it was before. Microsoft: As a major investor, we must prevent it from going solo. The relationship between Microsoft and OpenAI is simple: Microsoft fears OpenAI will leave, and OpenAI fears Microsoft's control. Now, having invested so much money, Microsoft's biggest concern is actually OpenAI's excessive success. Microsoft's early support for OpenAI was very substantial: money, computing power, cloud services, enterprise customers, access to Office, Windows, and GitHub—truly a major financial backer. OpenAI's early ability to train and deploy massive models was absolutely inseparable from the support of Microsoft Azure. After ChatGPT's explosive popularity, Microsoft quickly integrated OpenAI models into Copilot, Office, Bing, and enterprise services. It seems Microsoft has hit the jackpot; years ago, it was paying the price for slightly lagging behind in the mobile internet era, but in recent years, thanks to OpenAI, it has turned its fortunes around and returned to the center of the technological narrative. The more successful this relationship became, the more awkward it became for both sides. Microsoft wanted OpenAI to help it sell Azure, Office, and Copilot; OpenAI, on the other hand, wanted to build its own consumer portal, enterprise platform, developer ecosystem, search products, operating system distribution, and even consumer hardware. Therefore, Microsoft began to prepare contingency plans. On the one hand, it continued to provide OpenAI with computing power and channels; on the other hand, it trained its own models, introduced competitors like Anthropic, and reduced its dependence on OpenAI as a single vendor. OpenAI, similarly, was unwilling to completely hand over its lifeline to Microsoft. It began seeking more cloud service providers, partnering with Oracle, CoreWeave, and even Google Cloud, continuously eroding Azure's exclusivity. While both sides verbally emphasized their strategic cooperation, they were quietly preparing contingency plans to "get by without each other." Anthropic: Brothers Descend the Mountain, Who is the Rightful Successor? The feud between OpenAI and Anthropic is a classic case of internal strife. Many members of Anthropic's core founding team came from OpenAI, including Anthropic's co-founder and CEO, Dario Amodei. The reasons for their initial separation mainly revolved around AI safety, the speed of commercialization, and corporate governance. Simply put: Anthropic believed in greater caution, while OpenAI believed that products should be released and monetized first. So after the Anthropic team left, they established a very distinct brand image: we prioritize safety, explainability, and long-term risk, and we won't floor the gas pedal for growth. The subtext of this narrative is obvious—who's recklessly accelerating? Certainly not us. OpenAI naturally wouldn't admit to being the driver who recklessly accelerated, so the debate quickly escalated from a conceptual argument to product competition. ChatGPT vs. Claude; OpenAI API vs. Anthropic API; Codex vs. Claude Code; enterprise clients vs. enterprise clients; researchers vs. researchers; security narratives vs. security narratives. When both sides truly entered the same commercial arena, the initial ideological clashes quickly transformed into a pragmatic battle over revenue. Later, even financial terminology and revenue recognition methods became weapons in their attacks. It's like two martial arts brothers who, upon parting ways, were fiercely opposed, declaring their paths diverged and they would not work together. After years of fighting, they discovered that what they were truly fighting for was: who would inherit the sect, who would maintain the lineage, and who would claim the title of leader of the martial arts world. Musk and xAI: I Hate Myself for Founding You
If OpenAI's conflicts with other companies are somewhat driven by commercial interests, its relationship with Musk has taken on a strong personal grudge. As is well known, Musk is one of the co-founders of OpenAI.
If OpenAI's conflicts with other companies are somewhat driven by commercial interests, then its relationship with Musk has taken on a strong personal grudge.
As is well known, Musk is one of the co-founders of OpenAI.
So Musk began to strongly emphasize and attack OpenAI for deviating from its original intentions. Not content with just verbal attacks, he also founded xAI and launched Grok, directly engaging in a battle with OpenAI. From then on, the two sides entered a very stable relationship: OpenAI launched a product, Musk mocked it; OpenAI raised funds, Musk questioned it; OpenAI restructured, Musk sued; xAI releases model, Musk executives make sarcastic remarks; Employees of both sides exchange barbs again. The lawsuits between OpenAI and Musk are numerous enough to warrant a separate legal column. Coincidentally, xAI previously sued OpenAI over non-compete agreements and trade secrets, accusing former employees of providing Grok-related information to competitors. This makes Apple's new lawsuit particularly relevant. Not long ago, OpenAI emphasized in court that employees changing jobs and sharing past work experience did not mean the new employer stole trade secrets. Now Apple has turned around and is coming after them with more detailed documents, chat logs, and hardware allegations. OpenAI's lawyers haven't even had time to change their story. Finally: Why does OpenAI always manage to turn its partners into competitors? There's nothing new under the sun. More than two thousand years ago, ancient Greece had two of the most powerful city-states, Sparta and Athens. They even fought side-by-side against a common enemy and maintained a semblance of peace for a time. But as Athens grew stronger, the old hegemon Sparta became increasingly uneasy; and the more Sparta guarded against them, the more Athens felt oppressed. Ultimately, fear, suspicion, and conflicting interests accumulated, evolving into the Peloponnesian War, which lasted for many years. Historian Thucydides later wrote that the real cause of the war was "the growth of Athenian power and the fear that this growth aroused in Sparta." This situation has been summarized as the "Thucydides Trap": a rapidly rising power begins to encroach on the sphere of influence of the old hegemon; the old hegemon, out of fear and vigilance, continuously tightens its grip, and even if neither side initially intends war, they will gradually slide towards conflict. OpenAI is the rapidly expanding Athens of the AI era. Initially, it was just a large model laboratory, requiring funding and computing power from Microsoft, access to Apple devices, infrastructure from cloud providers, and the entire Silicon Valley ecosystem to support it. Back then, everyone was willing to cooperate with it, even though it was still in its infancy. It had potential, but wasn't yet powerful enough to threaten anyone's foundation. However, with the emergence of Chatgpt, OpenAI was no longer content with just building models. It gradually expanded into search, office software, programming tools, browsers, agents, operating system portals, and consumer hardware, redefining the way humans interact with computers. With each step it took, it encroached on the core territory of a tech giant. As a result, the companies that once welcomed, invested in, and supported it began to feel increasingly uneasy, like Spartans. Perhaps, an ultimate war is brewing between OpenAI and other AI upstarts and traditional internet tech giants.
Alex
Catherine
Clement
Jasper
Jixu
Hui Xin
YouQuan
Kikyo