John Sculley, who steered Apple through the 1980s and early 1990s, didn’t mince words at the Zeta Live conference in New York this week: OpenAI is the first company in “many decades” to pose a genuine competitive threat to Apple — and the battleground is artificial intelligence. The comment landed like a splash of cold water for anyone still picturing Apple as untouchable on the strength of design, hardware and an ironclad ecosystem.

Sculley’s thesis is simple and blunt. Apple, he said, has not historically leaned into AI the way newer tech giants have, and that gap is now visible in product cadence and public perception. He pointed to the company’s halting rollout of major AI-driven features — most notably a long-promised overhaul of Siri — as evidence that Apple’s approach has been slower and more cautious than rivals’.

A different philosophy, and different trade-offs

To understand why Sculley’s observation matters, you have to understand Apple’s playbook. For more than a decade, the company leaned into two intertwined ideas: doing a very small number of things extremely well, and protecting user privacy by keeping as much processing on-device as possible. That approach has real advantages — it differentiates Apple in an era of rampant data harvesting — but it also complicates rapid, cloud-first moves into large language models and other compute-hungry AI features that companies like OpenAI, Google and Microsoft have prioritized. Apple’s public financial and infrastructure moves show it’s pushing hard on AI — it has committed massive capital to U.S. investment and chips — but that doesn’t make feature rollouts instantaneous.

That cautious strategy is part of why some Apple engineers and executives have been open about delays. At WWDC and in later interviews, Apple executives acknowledged that the Siri upgrade didn’t meet the company’s internal quality thresholds and was pushed back for reliability reasons. In plain terms: Apple decided it would rather delay a product than ship something that feels flaky — and that conservatism looks, from the outside, like falling behind.

OpenAI isn’t just a new model vendor — it’s building hardware and buying design talent

Sculley’s specific naming of OpenAI as a competitor isn’t just about model performance. It’s about the company’s broader ambition. OpenAI has been moving beyond API services and chat windows: it acquired Jony Ive’s device startup earlier this year and is clearly aiming to make hardware — or at least hardware-adjacent devices — that put its models front-and-center for consumers. That combination of leading-edge AI and serious industrial design talent is the exact pressure point Sculley highlighted.

That marriage of design pedigree and AI muscle is precisely what worries old-guard companies that succeeded by marrying hardware and software. Sculley explicitly called out Jony Ive — the designer of the iMac, iPod, iPhone and iPad — as someone uniquely positioned to “bring that dimension” to OpenAI’s work with Sam Altman. If OpenAI can solve the hard problems of making AI feel delightful and trustworthy on a physical device, Apple’s longstanding advantage in design may cease to be a moat and instead become a battleground.

Reality check: building a consumer AI device is hard. Public reporting suggests OpenAI and Ive face technical and philosophical hurdles — from defining a device “personality” to ensuring privacy and managing heat, power and on-device compute requirements. These are nontrivial engineering problems, and critics caution that the hype may outpace deliverability. But even the attempt reshapes the competitive map.

The “agentic era” and the business model shift

Sculley used another phrase that’s gaining traction in boardrooms: the “agentic era.” The gist is that future interfaces will rely less on a thousand discrete apps and more on intelligent agents that accomplish tasks across applications for users. For companies that can change how value is packaged: instead of one-off purchases or ad-funded ecosystems, subscription and recurring revenue for always-on agents looks more natural. Sculley argued this will push more businesses toward subscription models — something Apple already uses heavily, but which could expand in both scope and intensity.

For Apple, the agentic future raises two practical questions. One: can Apple’s privacy-first, silicon-driven strategy support agentic experiences at scale? Two: can Apple translate those experiences into subscription offerings that don’t alienate users or regulators? The company is investing in chips and data centers, but translating investment into compelling, fast-moving product releases is the real test.

Leadership, optics and what comes next

Sculley also weighed in on leadership at Apple, noting public speculation about Tim Cook’s possible retirement and suggesting that whoever follows Cook must shepherd the company through the agentic transition. Leadership changes — whether imminent or not — always amplify questions about strategy. For a company whose brand is tightly tied to singular product visions, questions about pace and direction are now front-page material rather than trade press chatter.

So what should investors, users and Apple-watchers look for? Watch product cadences (especially Siri and Apple Intelligence rollouts), the company’s acquisitions and hiring in AI and device design, and whether Apple experiments with new subscription bundles anchored to agentic features. And watch OpenAI’s product trajectory too — if it can make devices that feel as well-built and thoughtfully designed as Apple’s, the market map could look very different within a few years. Reporting so far shows the path isn’t easy for OpenAI either, but the mere effort forces Apple to accelerate and rethink long-held assumptions.

Sculley’s comments are provocative because they reframe a familiar story: Apple’s dominance was never guaranteed — it was earned by design, tight integration and a long-term product mindset. The rise of large language models and an AI-first set of companies like OpenAI changes the rules of the game. For Apple, the choice is not whether to play — it already is — but how fast and how visibly it will change the parts of its approach that made it successful in the first place.

If you squint, this is a story about reinvention. Apple has reinvented itself before. But the clock is ticking differently now — the cadence of AI and the appetite of consumers (and regulators) for new experiences move faster than the company’s historically deliberate release schedule. Whether Apple’s next reinvention will be a smooth one or a fight for footing is the real story to watch.