Why does OpenAI always manage to turn partners into competitors?

Written by: Biteye core contributor Alan

Edited by: Biteye core contributor Denise

Rewind the clock two years, Apple had just welcomed ChatGPT with great fanfare into the iPhone at WWDC (Apple Worldwide Developers Conference).

The scene at that time was simply a match made in heaven: Apple boasts the largest and most extensive device empire in the world, while OpenAI was responsible for tutoring Siri, who had still not learned to speak human language. Apple lacked models, OpenAI lacked entry points, each side got what they needed, and they were a perfect match.

But today, Apple has turned around and sued OpenAI in court.

On July 10, 2026, Apple filed a lawsuit against OpenAI, OpenAI’s hardware company io Products, and two former Apple employees, Chang Liu and Tang Yew Tan, in the United States District Court for the Northern District of California. They accused them of systematically stealing Apple’s confidential documents, hardware components, manufacturing processes, and unpublished product information through recruitment, interviews, departure guidance, and supply chain cooperation to accelerate OpenAI's consumer hardware business.

Business competition is that straightforward.

What is called strategic cooperation is often just a dignified term for when both parties have not yet clashed head-on.

Once they start competing for the same users, talent, supply chains, and next-generation hardware entry, even the closest allies can turn against each other. Moreover, OpenAI's expansion over the past few years has almost stepped into the core territory of a tech giant with every move forward.

However, upon closer inspection, why does it seem like OpenAI has issues with everyone?

This company has a near-magical ability—it always manages to shake hands and cooperate with others first, then takes its business right to their doorstep, ultimately successfully developing partners into competitors, and turning investors into "family members" who are on guard against OpenAI.

This editor has briefly summarized the love-hate relationships between OpenAI and leading tech companies.

01 Apple: Once Sweet, Now in Court

The relationship between Apple and OpenAI is a classic case of a flash marriage and divorce in the AI circle.

By 2024, Apple had clearly fallen behind in generative AI. ChatGPT had already taken a stronghold in the global user mindset, Google was vigorously promoting Gemini, and Microsoft was cramming Copilot into Office to assist with office tasks. Meanwhile, Apple, as the "AI Big Brother" during the era of ChatGPT, was still struggling to prove on the developer conference stage that Siri could genuinely understand human language.

There was nothing to do but seek external help.

So Apple welcomed ChatGPT into Apple Intelligence. When Siri encountered questions it couldn’t answer, it could politely ask, "Would you like me to ask ChatGPT?"

OpenAI thus gained access to one of the world’s most valuable consumer electronics entry points; Apple, in turn, temporarily masked its lack of large model capabilities.

But in reality, the two companies did not want the same thing from the beginning.

Apple hoped ChatGPT would be a plugin—ideally residing quietly deep in the system, called upon when needed, and not stealing the spotlight when not. Apple controlled users, hardware, systems, and distribution, while OpenAI was responsible for providing a portion of the capability.

However, what OpenAI wanted was clearly more than just a high-end outsourcing solution living in the iPhone; it aimed to become the new entry point for users. If in the future users no longer open an app but instead directly tell the AI, "Get this done for me," then whoever controls the AI assistant could bypass traditional operating systems.

OpenAI later discovered that Apple was a treasure trove and brought in a team led by former Apple Chief Design Officer Jony Ive, one of the key figures behind products like the iPhone, iPad, and MacBook, making a grand entry into consumer hardware. Thus, Apple could no longer sit idle, failing to mention the past friendship of Siri and ChatGPT, now simultaneously introducing Gemini into the core capabilities of the new Siri and suing OpenAI.

According to Apple's lawsuit, it's no longer just a matter of "normal recruitment."

Apple claims that a former employee, Chang Liu, who left to join OpenAI, continued to access Apple's internal servers through vulnerabilities and downloaded a large amount of confidential engineering data; he was also accused of helping other Apple employees copy files and evade security reviews, preparing for OpenAI interviews. Apple further claims that OpenAI's recruitment system guides Apple employees on how to handle exit reviews, reminds them not to disclose their destination too early (especially not mentioning OpenAI), not to sign documents casually, and to extend their system access as long as possible.

Even during interviews, OpenAI's former Apple employees would use Apple’s internal project codenames, asking candidates to prepare "technical in-depth analyses" related to their current work. Some candidates were asked to bring Apple’s batteries, circuit boards, logic boards, and prototype components to the interview. Of course, as of now, these are still one-sided accusations from Apple, and whether they ultimately hold up will depend on evidence disclosure and court proceedings. But regardless of the lawsuit's outcome, the relationship between Apple and OpenAI is unlikely to return to its former state.

02 Microsoft: As a Financial Backer, Be Careful of Its Independence

The relationship between Microsoft and OpenAI is simple: Microsoft fears OpenAI will run away, and OpenAI fears Microsoft will control it. As of now, after Microsoft has invested so much money, its biggest concern is that OpenAI could become too successful.

In its early days, Microsoft provided OpenAI with very tangible resources: money, computing power, cloud services, enterprise customers, Office entry, Windows entry, GitHub entry—truly a financial backer.

OpenAI's ability to train and deploy ultra-large models early on absolutely benefited from Microsoft's Azure support. After the explosion of ChatGPT, Microsoft quickly integrated OpenAI's models into Copilot, Office, Bing, and enterprise services. Microsoft appears to have hit a jackpot, previously paying the price for trailing in the mobile internet, but in recent years, turning things around overnight thanks to OpenAI, positioned back in the center of technology narratives.

But the more successful this relationship becomes, the more awkward it becomes for both parties.

Microsoft wishes OpenAI would help sell Azure, Office, and Copilot; OpenAI, on the other hand, hopes to establish its own consumer entry, enterprise platform, developer ecosystem, search product, and operating system agent, even consumer hardware. Thus, Microsoft started to leave backdoors. On one hand, it continues to provide OpenAI with computing power and channels; on the other, it trains its own models, brings in competitors like Anthropic, and reduces reliance on OpenAI as a single supplier. OpenAI also does not wish to completely hand over its lifeline to Microsoft.

It began seeking more cloud service providers, collaborating with Oracle, CoreWeave, and even Google Cloud, continuously weakening Azure’s exclusive status. Both sides have always emphasized strategic cooperation verbally, but both are quietly preparing plans to "survive without each other."

03 Anthropic: Brothers Falling Out, Who Is the Legitimate Heir?

The conflict between OpenAI and Anthropic is a standard case of former allies turning against each other.

The core founding team of Anthropic includes a number of individuals from OpenAI, including Anthropic's Co-founder and CEO, Dario Amodei.

The reasons for their split initially revolved around AI safety, commercialization speed, and corporate governance. Simply put: Anthropic believed in being more cautious, while OpenAI thought products should be released to earn money.

After the Anthropic team left, they established a very distinctive brand image for themselves: they prioritize safety, interpretability, and long-term risks, and will not floor the accelerator for growth.

The underlying message of this narrative is also very clear—who is reckless? Certainly not us.

OpenAI, of course, would not admit to being that reckless driver, so the debate quickly escalated from ideological differences to product competition. ChatGPT against Claude; OpenAI API against Anthropic API; Codex against Claude Code; enterprise clients against enterprise clients; researchers against researchers; safety narratives against safety narratives. When both sides truly entered the same commercial arena, their previous ideological disputes quickly turned into the most pragmatic revenue competition. Later on, even financial metrics and income recognition methods could become weapons for mutual attacks.

It’s like two brothers turning on each other when they split, saying they don’t share the same path. After fighting for a few years, they finally realize that what they’re truly fighting over is: who controls the gates, who inherits the legacy, and who holds the title of the leader of the martial arts world.

04 Musk and xAI: I Hate That I Founded You

If the conflict between OpenAI and other companies still holds some commercial interest, the relationship with Musk is already heavily tinted with personal grievances. It is well known that Musk is one of the co-founders of OpenAI.

Back then, OpenAI still carried the halo of being "open," "non-profit," and "for all humanity," aiming to prevent advanced AI from being monopolized by a few tech giants. Later, Musk departed, and OpenAI gradually moved toward commercialization, becoming deeply bonded with Microsoft as its models became increasingly closed and valuation skyrocketed.

From Musk's perspective, it is akin to: the hopeful school they once helped establish together became a public-funded school that turned private years later, with a sign of Microsoft hanging at the gate.

Thus, Musk began to vocally criticize OpenAI for straying from its original intentions, and mere verbal attacks weren’t enough; he founded xAI, launched Grok, and directly entered into a rivalry with OpenAI.

Since then, the two parties have entered into a very stable relationship:

• OpenAI releases products, Musk mocks;
• OpenAI raises funds, Musk questions;
• OpenAI adjusts its structure, Musk sues;
• xAI releases models, Musk's executives throw out petty comments;

Employees from both sides volley insults across the void.

The court cases between OpenAI and Musk have accumulated to the point that they could fill an entire legal column.

Coincidentally, xAI has sued OpenAI over non-compete agreements and trade secrets, accusing former employees of bringing information related to Grok to competitors.

This makes Apple’s new lawsuit feel particularly authentic.

Not long ago, OpenAI even emphasized in court: employees switching jobs and sharing past work experiences do not imply that new employers have stolen trade secrets. Now, as Apple turns around with more detailed documents, chat logs, and hardware accusations, OpenAI’s lawyers are scrambling to come up with a different argument.

05 In Conclusion: Why Does OpenAI Always Manage to Turn Partners into Competitors?

Under the sun, there is truly nothing new.

Over two thousand years ago, there were two of the most powerful city-states in ancient Greece, Sparta and Athens. They even once stood side by side against a common enemy and maintained a period of superficial peace. But as Athens continued to grow stronger, the old hegemon Sparta began to feel increasingly uneasy; the more Sparta guarded against it, the more Athens felt it was being oppressed.

Ultimately, fear, suspicion, and conflicts of interest accumulated layer upon layer, evolving into the prolonged Peloponnesian War. The historian Thucydides later wrote that the true cause of the war was "the rise of Athenian power and the fear that growth instilled in Sparta."

Later generations summarized this situation as the "Thucydides Trap": a rising power rapidly ascends, beginning to encroach on the sphere of influence of an established hegemon; the old hegemon, due to fear and prevention, continuously tightens the space, and even if neither side initially intends to go to war, they will gradually slide into conflict.

And OpenAI is the Athens of the AI era, rapidly expanding. Initially, it was merely a large model laboratory, needing Microsoft’s funding and computing power, Apple’s device entry, cloud vendors’ infrastructure, and the entire Silicon Valley ecosystem to support it.

At that time, everyone was willing to cooperate with it, as it had enough potential but was not strong enough to threaten anyone’s foundation. However, with the emergence of ChatGPT, OpenAI was no longer satisfied with just being a model. It began to step into search, office software, programming tools, browsers, agents, operating system entry, and consumer hardware, redefining the way humans interact with computers.

With every step it takes forward, it steps into the core territory of a tech giant. Thus, companies that once welcomed, invested in, and supported it started to feel increasingly uneasy, much like Sparta. Perhaps the new AI elites represented by OpenAI and the traditional internet tech giants are still headed toward an ultimate battle.