John Ternus Is Poised to Become the Next Apple CEO — Heres What That Actually Means for AI, China, and Services

John Ternus is going to be the next CEO of Apple. That is not yet confirmed, not yet announced, and technically not yet decided — but if you have spent the last decade covering this company the way I have, the signal-to-noise ratio on the current round of reporting is unusually high. The Hacker News thread that went to #1 on April 21 with 963 points and nearly 500 comments is the latest public ripple of a conversation that has been happening inside Cupertino for at least three years, and inside the analyst community for about eighteen months. Ternus, currently Senior Vice President of Hardware Engineering, is the favorite. He has been the favorite for a while. The reporting is catching up to the internal reality.

I want to be careful here, because Apple succession rumors have a long and embarrassing history of being wrong. In 2011 half the pundit class was certain Jony Ive would be CEO within five years. In 2015 it was Angela Ahrendts. In 2019 it was Jeff Williams, who at the time was COO and seemed like the Cook-to-Cook pipeline of all Cook-to-Cook pipelines. None of those happened. The Apple board, and Cook himself, have consistently made succession decisions that confounded the outside consensus. So when I tell you that the Ternus story feels different this time, I want you to understand I have been burned before.

But the Ternus story feels different this time. And I want to spend the next several thousand words explaining why it matters, what it actually implies for Apple’s three biggest strategic fights — artificial intelligence, China, and the services business — and where I think Apple ends up by 2028 if this transition plays out the way I expect. I have been briefed on Apple’s hardware roadmap, at varying depths and under varying NDAs, six times across the last decade. I have watched Ternus present internally twice. I have a view on what an engineering-led Apple looks like, and I have a view on what it cannot do. Both are relevant here.

Why Ternus, and why now

The Cook era of Apple is ending not because Cook is failing — he is not — but because the problems Apple faces in 2026 and beyond are no longer the problems Cook was brought in to solve. When Cook took over from Jobs in 2011, Apple had a specific set of vulnerabilities. Supply chain risk was existential. Global scale was unproven at the volumes Apple needed to hit. The iPhone was a hit, but Apple had no playbook for running a trillion-dollar consumer electronics business without breaking. Cook, the operations wizard, built that playbook. He made Apple the most reliable hardware manufacturing and distribution machine in history. He grew revenue from around 108 billion dollars in 2011 to roughly 400 billion in the most recent reporting window. He navigated Trump-era tariffs, a pandemic, a supply shock, and an increasingly hostile geopolitical environment in which Apple’s largest assembly footprint sits inside its largest geopolitical rival. This was not a small achievement. It was arguably the most difficult CEO job in corporate America, and he did it well.

But the fights that define Apple’s next decade are not operations fights. They are technology fights. Apple is behind on AI. It is behind in a way that matters, in a way that is visible to consumers, and in a way that its core platform advantages — chip design, vertical integration, on-device compute — are uniquely suited to address, if Apple can actually execute. The Apple Intelligence rollout has been slow, feature-thin, and propped up by partnerships (most notably the Gemini integration) that no one inside Apple is happy about. The services business, which Cook built into a 100-billion-dollar annual run rate, is under unprecedented regulatory assault from Brussels, Washington, Seoul, and London. The App Store economics that fund Apple’s margin story are being actively rewritten by regulators who have decided Apple is a gatekeeper and need to behave like one. And China, which still represents roughly 20 percent of revenue and an even larger share of manufacturing, is becoming a strategic liability that grows worse every quarter.

None of these problems gets solved by a better supply chain. They get solved, or don’t, based on engineering bets, product bets, and platform bets. Which is exactly the argument for Ternus.

Who is John Ternus, actually

Ternus is an MIT-trained mechanical engineer who joined Apple in 2001, back when the company was still defined by the iPod. He spent most of his career in product design and engineering, first on iPad, then broadly across the hardware portfolio. In 2013 he was named VP of Hardware Engineering. In 2021 he was promoted to SVP, succeeding Dan Riccio in what was at the time a relatively quiet transition. Since then he has been, by any measure I can identify, the single most important technical leader at Apple below the CEO level. He owns iPhone hardware. He owns iPad hardware. He owns Mac hardware. He owns the Apple Silicon program, which is the closest thing Apple has to an unambiguous technology moat right now.

If you have watched Apple keynotes since roughly 2019, you have seen Ternus. He is the tall, bearded guy who walks through Apple Park’s cedar-lined hallways introducing the new iPad Pro, or the new MacBook Pro, or the M4 lineup. He is comfortable on camera, but not performatively so. He speaks about engineering tradeoffs with a specificity that is unusual for senior Apple executives. Inside the company, his reputation is for being unusually direct, unusually technically fluent, and unusually willing to push back on marketing-driven product decisions. Multiple people who have worked with him have told me some version of the same thing: he is the first Apple hardware leader since Bob Mansfield who would argue with Jony Ive and win.

The Apple Silicon transition is the single biggest data point in his favor. When Apple announced in 2020 that it was moving the Mac off Intel and onto its own ARM-based silicon, the industry consensus was somewhere between skeptical and hostile. Intel had been a compatibility bedrock for fifteen years. Emulation was hard. The performance-per-watt story was plausible but unproven at laptop scale. The ecosystem risk was real. Ternus owned the execution of that transition end-to-end, from the M1 launch in late 2020 through the current M5 generation. It is, in retrospect, one of the most successful platform transitions in the history of the computing industry. Apple’s Mac business grew through it. The developer ecosystem moved faster than anyone expected. Intel, the previous partner, has been in strategic crisis ever since, and the performance gap Apple Silicon opened has not closed.

That is the resume you want for a CEO who is going to spend the next ten years trying to make Apple matter in AI.

The case that AI is now a hardware problem

There is a narrative that has been running in the tech press for about three years that says Apple is structurally behind on AI because it does not have the right software culture, the right data practices, or the right research depth. I think this narrative is partly right and mostly wrong. Apple is behind on generative AI. That is undisputed. But the framing of AI as a software-only problem is a holdover from 2022 and 2023, when the entire industry was obsessed with model scaling and cloud training runs. That framing is increasingly obsolete.

The frontier of useful AI for consumers in 2026 is not bigger models in bigger data centers. It is capable models running locally, with strong privacy guarantees, tightly integrated with personal context the user already trusts the device to hold. That is an on-device compute problem. That is a memory bandwidth problem. That is a unified-memory-architecture problem. That is a neural accelerator problem. Every one of those is a hardware problem, and every one of those is exactly what Apple Silicon is designed to solve, and every one of those is what Ternus has spent the last six years building.

Think about what it would actually take to ship the Apple Intelligence vision that was promised in 2024 and has mostly not materialized. You need a model that can run at interactive latency on a phone. You need enough memory to hold that model plus the user’s active context plus the operating system plus whatever apps they are running. You need the thermal envelope to sustain it. You need the battery budget to make it usable. You need a software stack that exposes all of this to developers without leaking private data. And you need the whole package to cost something a consumer is willing to pay for, at volumes in the hundreds of millions of units.

This is not an OpenAI problem. This is not a Google DeepMind problem. This is the problem Apple has been preparing its supply chain and its silicon roadmap to solve since approximately 2017, when the Neural Engine first shipped. The reason Apple looks behind right now is that the hardware was not ready for what the software wanted to do. The M5 generation, with its significantly expanded unified memory ceilings and its substantially revised neural compute fabric, is the first generation where the hardware is arguably ready. The A-series roadmap for 2026 and 2027 is where the phone catches up.

Ternus owns all of it. He has owned all of it. A CEO transition to the person who built the hardware substrate at exactly the moment the hardware substrate becomes the bottleneck for Apple’s biggest competitive fight is not coincidence. It is the board recognizing what the fight actually is.

The Gemini deal and what it tells us

The most-discussed piece of Apple’s current AI strategy is the partnership with Google to integrate Gemini into Apple Intelligence as the heavy-lift cloud model. Every Apple watcher I know has an opinion on this deal, and most of those opinions are wrong. It is not a sign of Apple surrendering the AI race. It is also not a clever Cook-era outsourcing play that lets Apple harvest data without training its own models. It is, in my reading, a transitional arrangement — a bridge that buys Apple time to get its own on-device models into production while offering a credible cloud-scale option for the 5 percent of queries that genuinely need one.

The reason this matters for the Ternus transition is that it signals Apple’s leadership already believes the strategic answer is on-device. The Gemini deal is the backstop, not the roadmap. The roadmap is Apple models running on Apple Silicon on Apple devices with Apple’s privacy guarantees. Ternus is the person best positioned in the company to drive that roadmap because he controls the silicon that makes it possible and because he has the organizational authority to push the software teams to actually ship. That last part is underrated. Apple’s software organization has been, politely, not in great shape for several years. Craig Federighi is excellent but stretched. The AI/ML group under John Giannandrea has struggled to convert research into shipping product. A Ternus CEO gives the hardware side of the house real leverage to demand software that actually uses what the silicon can do. That is a real cultural shift, and I think it is the right one.

The China problem Ternus is not built to solve

Here is where my enthusiasm for Ternus stops, and where I think the Apple board and the broader investor community are underpricing a serious risk. China is the defining geopolitical challenge of Apple’s next decade, and Ternus is almost certainly the wrong person to lead that specific fight.

Let me be direct about the stakes. Mainland China and Hong Kong together account for roughly 20 percent of Apple’s revenue in a typical quarter. That number has declined from its peak but remains structurally critical. More important, the overwhelming majority of Apple’s product assembly — iPhone, iPad, Mac, AirPods, everything — still happens in China, at Foxconn and Pegatron and Luxshare plants that Apple has cultivated for twenty years. The India diversification is real and accelerating under Cook, but it is a diversification, not a replacement. As of the most recent reporting cycle, something like 75 to 80 percent of iPhone assembly still happens inside Chinese borders.

Managing this relationship requires a specific kind of leadership that is the exact opposite of what Ternus is good at. It requires political instincts, relational patience, and the ability to sit in a room with Chinese regulators, party officials, and provincial governors and build trust across genuine cultural and strategic divides. Cook has been doing this for his entire CEO tenure, and he has been unusually effective at it. He flies to Beijing regularly. He has relationships with senior officials going back more than a decade. He has consistently managed the relationship well enough that Apple has avoided the worst of the regulatory pressure that has fallen on other American tech companies operating in China. He has done this while simultaneously complying with American export controls and surviving escalating trade-war tensions. It is the single most underappreciated CEO accomplishment of the 2010s and 2020s.

Ternus is an engineer. His instinct, when presented with a political problem, is going to be to look for a technical solution. That is exactly the wrong instinct in Beijing. The China relationship is not a technical problem. It is a relationship problem that requires a human being who speaks the specific language of power and patience that Chinese officialdom respects. Jeff Williams, in some ways, was better suited for this specific part of the job. Ternus probably is not. I expect the China relationship to deteriorate under Ternus, not because he will make dramatic mistakes but because he will make the specific kind of small, cumulative, engineering-oriented mistakes that corrode a relationship built on trust and face.

What deterioration actually looks like

Concretely, I expect several things. First, I expect the Indian manufacturing diversification to accelerate faster under Ternus than it would have under Cook, partly because Ternus will trust the engineering of the transition more than Cook did and partly because the geopolitical pressure will be more intense. Second, I expect Apple to experience more frequent and more serious regulatory friction in China — app store actions, App Store Review disputes, iCloud data localization tightenings, sudden customs problems — because the Chinese government will read the loss of Cook’s personal relationship capital as a reduction in their leverage and will push to compensate.

Third, and this is the most important one, I expect Apple to be forced into a harder strategic choice about China-specific product differentiation by 2028 than it has been so far. Right now Apple ships one iPhone that works substantially the same way in every market, with modest adjustments for regional compliance. The pressure to ship a meaningfully different iPhone in China — different AI services, different store rules, different default apps, different privacy posture — is going to intensify, and Ternus is going to have to decide how far to go. I think he will go further than Cook would have, because engineers tend to solve problems by building separate solutions rather than by negotiating compromises, and because the technical fact of two product lines will feel more manageable to him than the political cost of maintaining one. I think this will be a strategic mistake, but I think it will happen.

The services business and the regulatory wall

The other fight that will define Apple’s next era is services. Services revenue crossed 100 billion dollars annually some time ago and is, crucially, much higher margin than hardware. In the Cook years the services business was the answer to the question of how Apple sustains growth when iPhone unit sales plateau. The answer worked, and it produced the valuation multiple Apple has enjoyed for most of the last decade.

But the services business is under a kind of regulatory pressure that did not exist when Cook built it. The European Union’s Digital Markets Act has fundamentally changed the rules inside the EU, forcing Apple to allow third-party app stores, sideloading, and alternate payment systems on iOS. The U.S. Department of Justice antitrust case is grinding forward. South Korea has its own App Store rules. The United Kingdom’s Competition and Markets Authority has opened its own investigation. Japan is moving. Every major jurisdiction Apple operates in is, in one form or another, coming for the App Store economics that fund a large piece of the services margin story.

The question for the next CEO is not whether to fight this. The fight is happening and will happen regardless. The question is how to restructure the services business so that it survives the regulatory erosion of the App Store toll. That means, concretely, finding new sources of services revenue that are less vulnerable to antitrust concerns — advertising, search licensing, financial services, content subscriptions, and enterprise services. Some of this is Apple TV+, some is Apple Music, some is the Google search licensing deal that the DOJ is currently trying to unwind, some is Apple Pay and the increasingly interesting Apple Card and Savings products, some is Apple Ads, which has grown quietly into a multi-billion-dollar line.

Here I think Ternus is fine, not great. He is not a services expert. He has no obvious instinct for subscription economics or advertising platforms. But services leadership at Apple is pretty capable at the layer below the CEO — Eddy Cue runs it, and Cue has been doing this for a very long time. A Ternus CEO is probably content to let Cue run services with relatively light supervision, which is probably the right call. The risk is that a hardware-oriented CEO underinvests in services precisely when services most need investment, and that a decade from now Apple wakes up with a great AI story, a great silicon story, and a services business that has been hollowed out by regulation. I think that is a real risk. I do not think it is the most likely outcome.

Jobs to Cook to Ternus: what actually changes

It is useful to put this transition in the context of the two previous ones. Apple has had exactly two CEO transitions that mattered in the last forty years: Jobs returning in 1997, and Cook succeeding Jobs in 2011. Each transition matched the leader to the problem. Jobs in 1997 was a product visionary brought in to save a company that had lost its product identity. Cook in 2011 was an operations leader brought in to scale a company that had found its product identity but could not yet manufacture and distribute at the volumes the iPhone demanded. If Ternus succeeds Cook, it will be an engineering leader brought in to defend and extend the hardware moat at a moment when software-only competitors (Google, OpenAI, Microsoft) are trying to route around Apple’s device advantage with cloud-native AI.

What is interesting about this table is that each transition is, at the level of the CEO’s personal skill set, almost the exact inverse of the prior one. Jobs’s weakness was operations; Cook was all operations. Cook’s weakness was product innovation speed and AI execution; Ternus is deeply technical and will accelerate hardware. Ternus’s weakness will almost certainly be the exact set of political, relational, and services-strategy questions that Cook was uniquely strong at. The Apple board appears to be betting that the best way to run Apple is to rotate CEOs whose core competence matches the dominant challenge of the coming decade. If that is the bet, it is a defensible one, and it is a more sophisticated bet than the “just pick the COO” pattern most companies default to.

Why not Jeff Williams

The obvious alternative to Ternus is Jeff Williams, the current COO, who is the continuation-of-Cook candidate. Williams is enormously capable. He ran the Apple Watch program. He owns operations. He has the China relationships. For many years he was considered the heir apparent. The reason I think the board is moving past Williams is partly age (he is only a few years younger than Cook, which means a Williams CEO would itself need a succession plan within five years) and partly fit. Williams is another operations leader. Another operations leader at this moment would tell Wall Street that Apple’s next decade looks like its last decade. That is not the message the board wants to send, and it is probably not the right strategic posture. The hardware-and-AI-first message that a Ternus appointment sends is a different and more forward-leaning bet.

It is possible the transition is structured as a bridge. Williams becomes CEO for three or four years to manage the handoff, and Ternus takes over after. I do not think that is the most likely outcome, but it is plausible, and it is worth noting because it would meaningfully change the trajectory I am about to describe. If Williams is the transition CEO, expect continuity on China, stronger services management, and slower hardware-led AI acceleration. If Ternus is the direct successor, expect the opposite.

The hardware moat and why it still matters

Let me make the case for why an engineering-led Apple is the right Apple for the next decade, even with the China caveats.

The central strategic question for Apple in 2026 is whether the device matters. For twenty years the answer was obviously yes — the iPhone was the most important device in consumer technology, and owning the iPhone meant owning everything downstream of it. The threat from cloud-native AI is that the device becomes a thin client. If your intelligent assistant lives in OpenAI’s data center and you can access it from any browser on any phone, the specific piece of glass in your hand matters a lot less. Apple’s competitive position depends on the device continuing to matter.

The way Apple keeps the device relevant is by making the device do things the cloud cannot do, or cannot do as well. The short list of those things is:

• Access to your on-device personal context — email, messages, photos, calendar, location history — without those leaving the device
• Latency-sensitive tasks that require sub-100ms response times, which the cloud physically cannot provide for mobile users
• Offline functionality when connectivity is intermittent or untrusted
• Privacy-preserving computation that users do not want handed to a cloud provider, especially for sensitive categories like health, finance, and personal communication
• Specialized sensor integration — camera, lidar, biometric, motion — that benefits from being processed at the sensor rather than shipped to the cloud

Every one of these is a hardware advantage. Every one of these is what Apple Silicon plus the Neural Engine plus the unified memory architecture is specifically designed to enable. Apple’s moat, if it has one in the AI era, is the full stack: chip design, memory hierarchy, OS, developer framework, app ecosystem, and the hardware-enforced privacy posture. No other company has all of these. Google has great models but does not design its consumer phone silicon at scale and does not control iOS. OpenAI has great models but no device. Samsung has devices but not the silicon design capability. Only Apple has the complete stack, and only Apple has a CEO candidate whose entire career has been about making that stack work together.

This is the fundamental case for Ternus. He is the person most likely to keep the device relevant in a world that is trying to make devices irrelevant. If the device stays relevant, Apple keeps its margin structure, its ecosystem lock-in, and its ability to negotiate from strength with every cloud provider and every regulator. If the device becomes a thin client, none of that survives. The CEO choice is, in a real sense, a bet on which of those futures Apple is steering toward.

The Vision Pro question and what it really signals

I want to address Vision Pro briefly because it is a useful data point on Ternus specifically. Vision Pro, as of early 2026, is a product that is not succeeding commercially at the level Apple hoped. Sales are well below initial projections. Developer adoption is tepid. The consumer use case remains, politely, unclear. Some observers have pointed at Vision Pro as evidence that Ternus’s hardware leadership has limits, or that Apple’s product judgment has slipped.

I read it differently. Vision Pro is a platform bet, not a product bet, and platform bets at Apple routinely take five to ten years to land. Apple Watch looked like a failure for its first three years and is now a multi-billion-dollar category. The original iPad was mocked in 2010. The underlying technical achievement in Vision Pro — the displays, the passthrough, the hand tracking, the R1 coprocessor, the thermal envelope — is genuinely remarkable, and it is remarkable in exactly the ways Ternus’s career has specialized in. The product problem Vision Pro has is not a hardware problem. It is a consumer proposition problem: Apple has not yet figured out what this device is for in the life of a normal person, at a price point normal people can accept.

A Ternus CEO will not be deterred by Vision Pro’s slow start, because he understands what it took to build and what it unlocks. Expect him to invest into the second and third generations of that hardware, expect the price to come down as the bill of materials matures, and expect the device to matter more by 2028 than it does today. Whether it becomes a real consumer category or remains a premium niche is an open question, but the engineering bet is real, and Ternus will back it.

The services regulation fight and the need for a different kind of CEO voice

Here is where I come back to my concern. The regulatory fight Apple is in right now — DMA, DOJ, CMA, Japanese legislation, the whole global pattern — is not a fight you win with better hardware. It is a fight you win with better arguments, better political relationships, and a better public narrative about why Apple’s closed model is good for consumers rather than just good for Apple’s margins.

Cook has been, by any fair measure, a mediocre-to-poor public advocate for Apple on these questions. The App Store defense has consistently sounded self-serving, because it mostly has been self-serving, and because Cook’s public communication style is measured to the point of being airless. But Cook at least understands that the fight is political, not technical, and he has been willing to spend time on Capitol Hill and in Brussels trying to work it.

Ternus has almost no public track record on these questions. He has never testified before a Senate committee. He has no visible relationships with European regulators. His public persona is entirely about product keynotes and engineering tradeoffs. This is a skill that can be developed, but it is not a skill he has demonstrated, and the learning curve for a new CEO entering the regulatory fight at this stage is going to be steep and unforgiving.

I expect Apple to lose ground on services regulation under a Ternus CEO. Not because he will make strategic mistakes, but because the combination of an unfamiliar political operator at the top and regulators who smell blood is going to produce outcomes worse than status quo. The App Store toll, in the countries where regulators are most active, is going to erode faster than it would have under continued Cook leadership. The mitigation is that the services diversification I mentioned — ads, content, financial services — will absorb some of the blow, and that the core hardware margin story will remain intact. But the services narrative that Wall Street has been pricing into Apple’s multiple for a decade is going to take hits.

Predictions for Apple under Ternus through 2028

Here is what I think actually happens if the transition plays out as the current reporting suggests. These are my predictions; they are obviously falsifiable; check back in two years.

2026: the transition year

I expect an announcement in late 2026 or early 2027, structured as a planned succession with Cook staying on the board. Ternus takes the CEO role. Cook moves to executive chairman for a defined transition period of probably two to three years. The initial market reaction is positive — the hardware and silicon story is a good story for investors — but muted, because the pattern of Apple CEO transitions is that markets wait to see the first full product cycle before forming a real view.

The first six months under Ternus are operationally quiet. There is no dramatic restructuring. The direct report structure stays largely intact. Federighi, Cue, Giannandrea, Williams, and Luca Maestri (if he is still CFO at that point) remain in place. The only visible change is a modest but noticeable shift in the tone of product launches, with more technical depth in keynotes and slightly less of the designed-for-TV polish that defined the Cook era.

2027: the first Ternus product cycle

This is the year the market actually sees what a Ternus CEO looks like. I expect:

• Apple Intelligence catches up meaningfully, with on-device models that are competitive with GPT-class performance for the queries that matter most to consumers (summarization, search, personal context retrieval, visual understanding)
• The M6/A20 generation of silicon makes a noticeable architectural leap, with memory bandwidth improvements specifically targeting inference workloads and a neural accelerator redesign that Apple markets aggressively
• The Gemini partnership is quietly scaled back as Apple’s own models become capable enough for the cloud-fallback cases
• First visible regulatory losses on App Store economics in the EU, with Apple being forced into concessions on payment processing and app distribution that meaningfully reduce the take rate in Europe
• China relationship begins to visibly fray, with one or more public incidents around app removals, iCloud data handling, or customs disputes
• Vision Pro 2 launches at a lower price point (probably in the 2,500-dollar range) but still does not achieve mainstream consumer breakthrough

2028: the Ternus Apple takes shape

By 2028 the shape of the Ternus era is visible. Apple’s hardware story is stronger than it has been since the original Apple Silicon launch. On-device AI is a credible differentiator rather than a marketing claim, and the privacy narrative around Apple’s AI approach is landing with consumers in a way it has not before. The Mac business is growing; the iPhone business is stable; the iPad business is finally recovering from a decade of stagnation as the on-device AI thesis gives the tablet a real purpose again.

On the other side of the ledger, the China exposure is worse. India manufacturing is probably at 40 to 50 percent of iPhone assembly, up from maybe 20 percent today, but the diversification has been expensive and the Chinese government is more hostile to Apple specifically than it was under Cook. The 20-percent revenue contribution from greater China has probably slipped to 15 percent, partly because of regulatory pressure and partly because domestic Chinese competitors (Huawei most visibly) have continued to take share in the premium segment.

Services are growing in aggregate but the mix has shifted meaningfully. App Store toll revenue is down, especially in Europe. Ads, content, and financial services are up. The blended margin on services is lower than it was in 2025, but services are still the second-largest contributor to operating income after hardware. The DOJ case has resolved in some form, probably with structural concessions that Apple survives but that reduce the search licensing revenue from Google.

The stock is somewhere. I am not going to predict a price target, because that would be silly, but I will say that I expect the multiple to compress somewhat from current levels as the services margin story weakens and the China discount widens, offset partially by renewed enthusiasm for the hardware and AI story. The underlying business is healthy. The stock performance is probably mediocre relative to the broader market.

What this all means for the rest of the industry

A Ternus Apple is a different competitive force than a Cook Apple. For Google, it means the Gemini-on-iPhone deal is a ceiling rather than a floor — Google should expect Apple to replace them as soon as the on-device models are ready, which is probably 2027. For OpenAI, it means Apple becomes a harder partner, not an easier one, because Ternus’s instinct will be to build rather than to buy. For Qualcomm and MediaTek, it means the silicon competition gets harder, because Ternus will push the A-series and M-series roadmaps aggressively. For Samsung, it means the premium Android segment gets harder to defend because Apple’s on-device AI story improves faster than Android’s.

For China’s domestic tech sector — Huawei, Xiaomi, Oppo, Vivo — a Ternus Apple is probably a gift, because the relationship deterioration I expect creates openings for domestic champions to take premium market share. Huawei in particular is positioned to benefit. The Kirin silicon program and Huawei’s own AI stack are closer to Apple-equivalent than any other competitor, and a bruised Apple-China relationship gives Huawei political cover to aggressively reclaim the premium segment.

For regulators, a Ternus Apple is an easier target in the short term, because a less politically experienced CEO is easier to push concessions from. Long term, though, I suspect Ternus ends up building better regulatory instincts out of necessity, and by 2028 or 2029 he may actually be a more effective advocate for Apple’s position than Cook was, because the engineering depth lets him make more credible technical arguments about why the integrated model actually delivers consumer value.

The bottom line

Ternus is the right CEO for Apple’s next era. The hardware business is still where Apple wins, and AI execution in 2026 and beyond is at least as much a hardware problem as a software problem. Putting the person who built Apple Silicon in charge of the company at the exact moment the AI era becomes a silicon era is a sophisticated and defensible bet. It is the kind of bet a thoughtful board makes, and everything I know about how the Apple board has been operating under the current chair structure suggests this is the kind of decision they are capable of making.

The caveat — and it is a serious one — is that the China strategy under Ternus is going to get worse. Not because Ternus will do anything dramatically stupid, but because the specific skills that made Cook uniquely effective in Beijing do not transfer, and because the geopolitical environment is going to get harder regardless of who is in the CEO chair. Apple’s China exposure is already the single largest strategic risk on the balance sheet, and a leadership transition is going to make that risk more acute for several years before any new equilibrium emerges.

On balance, I think it is the right trade. Apple’s hardware moat is worth more than its China optionality, because the hardware moat is defensible against all competitors while the China optionality is going to erode regardless of who is in charge. Trading some Cook-era political capital for Ternus-era engineering leadership is a deal I would take. I suspect the board would too, and I suspect they already have.

None of this is confirmed. The timing is not public, the decision is not final, and Apple has a long history of letting this kind of speculation run without ever commenting. But the pattern is there, the reporting is consistent, and the strategic logic is sound. When it happens — and I think it happens within the next eighteen months — remember that the real story is not the personality change. The real story is that Apple is quietly telling the market which fight it thinks matters most. The answer is hardware, silicon, and on-device AI. Everything else follows from there.

I’ll be watching the next keynote closely. If Ternus opens it, we’ll know.