John Ternus takes the Apple CEO role in September with a hardware empire and a broken AI story to fix

John Ternus, Apple's senior vice president of hardware engineering, will become CEO on September 1, with Tim Cook moving to executive chairman. Ternus faces the challenge of fixing Siri, which has fallen behind ChatGPT, Gemini, and Claude, as Apple relies on Google's AI technology for its foundation models in a multi-year partnership reportedly worth around $1 billion a year.

John Ternus takes over Apple on September 1 with the hardware credentials investors understand and the AI problem users can already feel. Apple made the succession official on April 20: Tim Cook will move to executive chairman, and John Ternus, the company's senior vice president of hardware engineering, will become CEO on September 1. That is an orderly handoff by Apple standards. It still lands at an awkward moment. Ternus has spent 25 years inside Apple, joined the company in 2001, and holds a mechanical engineering degree from the University of Pennsylvania, according to Apple's own announcement. He knows the machines. Now he has to fix the assistant. The choice tells you what Apple wants to protect first. Ternus is not an outsider brought in to turn the company upside down. He is the hardware system made human: Mac, iPad, iPhone, AirPods, Apple Watch, Vision Pro, Apple Silicon, the long product roadmaps that Apple likes to keep quiet until the stage lights come on. Bloomberg reported in April that he had been viewed as Cook's heir apparent, while The New York Times had described him in January as the front-runner in Apple's succession planning. The market did not wake up to a mystery candidate. That is the comfort. It is also the risk. Morgan Stanley analysts wrote after the announcement that promoting Ternus showed Apple's emphasis on product at the center of its flywheel would remain, according to reports that cited the firm's note. Rosenblatt put the same idea more plainly, saying Apple was leaning into its hardware successes. Both readings make sense. Neither removes the hard part in front of him. Siri has been made to look old by ChatGPT, Gemini and Claude, and users do not grade Apple on succession planning. They ask the phone a question and know, in five seconds, whether the answer is good. The Siri Test Apple's answer is not to become OpenAI with better packaging. At WWDC in June, Apple presented Siri AI as a ground-up rebuild powered by the next generation of Apple Foundation Models, with requests handled on device where possible and more complex work routed through Private Cloud Compute. Apple's machine learning team said the new model family was custom built in collaboration with Google, and Apple's developer sessions showed Private Cloud Compute giving apps access to larger server models with privacy protections built into the system. That is a real thesis. Privacy is not decoration for Apple. It is one of the few promises the company has trained users to believe, and a phone-based assistant that can read personal context without spraying it across the open cloud is a serious product idea. You should want Apple to try that version of AI. The industry has enough chat boxes already. But here is the thing: the model still has to work. Apple confirmed in January that it would use Google's AI technology as a foundation for Apple Foundation Models, with TechCrunch reporting that earlier accounts put the deal at around $1 billion a year, though neither company confirmed the price. Ars Technica also reported that the Gemini-backed Siri deal was part of a multi-year partnership. That is not a footnote. It is Apple paying a rival for the intelligence layer inside one of its most visible features while it keeps the privacy architecture and product experience under its own control. That may be the right compromise. It is still a compromise. What Ternus Can Actually Fix Ternus has not run a public company before. He has not had Cook's years of managing earnings calls, Washington pressure, China supply chain risk, antitrust fights and investor nerves. What he has done is ship difficult hardware at scale. Apple's own release says he became vice president of hardware engineering in 2013, took over iPhone hardware in 2020 and became senior vice president of hardware engineering in 2021. If Apple's AI future depends on squeezing useful intelligence out of chips, batteries, sensors and operating systems working together, he is not the wrong person for the job. The bear case is simpler and nastier. On-device processing helps with privacy and latency on narrow tasks. It does not automatically produce an assistant that understands messy requests or handles long context - or one that knows what is happening outside your phone and replies like something more capable than a search field. ChatGPT and Gemini have already reset what users expect from software. A private Siri that still feels behind will not get credit for being private. It will just feel behind. CNBC reported in December, before the CEO announcement, that Deepwater Asset Management's Gene Munster said Apple had more time than people realized, but that the new Siri needed to be a "10 out of 10" when it arrived. That line has aged into Ternus's first operating test. Not because one analyst said it. Because Apple has already delayed the promised Siri overhaul once, brought in Google's model work, and used WWDC 2026 to set a new bar in public. For investors, the question is not whether the succession was tidy. It was. The question is whether Apple's hardware-first theory of AI can become a product people use every day, or whether it is a holding pattern while OpenAI and Google keep training users to look elsewhere first - and Anthropic waits in the queue. Ternus officially starts on September 1. Siri AI is due in the same fall window Apple just spent June selling to developers. There is not much runway between the handoff and the proof. Also read: Managing AI agents is now the most valuable engineering skill you can have https://startupfortune.com/managing-ai-agents-is-now-the-most-valuable-engineering-skill-you-can-have/ ; The AI industry is repricing itself around intelligence per dollar and Amazon is showing how https://startupfortune.com/the-ai-industry-is-repricing-itself-around-intelligence-per-dollar-and-amazon-is-showing-how/ ; Ford's $1 Billion DRAM Bill Shows the AI Boom Is Now a Car-Buyer Problem https://startupfortune.com/fords-1-billion-dram-bill-shows-the-ai-boom-is-now-a-car-buyer-problem/