Apple puts its future in AI in the hands of John Ternus, its new CEO

Apple is closing a historic chapter. Tim Cook will step down as chief executive officer on September 1, 2026, and will be replaced by John Ternus, until now head of hardware engineering. Cook remains tied to the company, but operational leadership changes hands after nearly 15 years.

Ternus takes over a company valued at around $4.5 trillion and, at the same time, the problem Cook never managed to fully solve: Apple arrived late to generative artificial intelligence and today relies on a third-party model to power its own assistant.

When Cook took over in August 2011, the question was whether Apple could survive without Steve Jobs. The market responded with caution: the stock fell 5.2% after the transition was announced, and Jobs passed away six weeks later.

The track record of the Cook era is resounding. Apple was worth around $350 billion and today hovers around $4.5 trillion, after briefly surpassing $5 trillion in July. It was the first company to hit one trillion in 2018, two trillion in 2020, and three trillion in 2022.

The other major achievement was structural: turning hardware into the gateway to a recurring services business. The App Store, Apple Music, iCloud, and Apple Pay went from generating around $9.4 billion annually in 2011 to exceeding $109 billion last year, with 1.5 billion active subscriptions spread across 2.5 billion devices.

In its latest fiscal quarter, Apple reported revenue of $109.4 billion: virtually the same amount it generated in the entire year of 2011. That figure alone illustrates the sheer scale of the transformation.

Since 2013, the company has dedicated more than $841 billion to share buybacks and has returned over a trillion dollars to its investors through buybacks and dividends combined. Cook was not a disruptive product CEO in the mold of Jobs: he was a CEO of supply chains, margins, and capital allocation, and in that arena, he had no equal.

The new CEO inherits an extraordinary financial position and an uncomfortable technological lag. The deep overhaul of Siri announced in 2024 has accumulated two years of delays, and Apple ultimately turned to Google's Gemini model to rebuild its assistant.

The move is more significant than it seems. Apple built its brand on vertical control: its chips, its operating system, its silicon, its supply chain. Supporting Siri's conversational layer with a third-party model breaks that principle in the technology that is redefining the relationship between user and device.

Apple's stated bet remains on-device AI: local inference on the Neural Engine, Private Cloud Compute for what doesn't fit on the phone, and privacy as a differentiating selling point. It is a strategy consistent with its history, but it demands exactly the profile that Ternus represents: someone who understands silicon, not just software.

Ternus is 50 years old, the same age Cook was when he took over. He joined Apple in 2001, and his career is closely tied to the products that defined the company's last two decades: the iPad, AirPods, and, above all, the Mac transition to in-house processors.

That last point carries the most weight. The leap from Intel to Apple Silicon in 2020 was the riskiest engineering operation of the Cook era, and it turned out flawlessly: better performance per watt, total control of the roadmap, and structural cost savings. Ternus was at the center of that move.

A mechanical engineer by training, he worked alongside Steve Jobs and later developed a mentoring relationship with Cook, who defines him with a phrase that summarizes the chosen profile: “He has the mind of an engineer and the soul of an innovator.” Within Apple, he is known for being discreet and meticulous, with a limited public presence outside of keynotes.

Ternus' real debut will not be a letter to shareholders; it will be a product. The unveiling of the iPhone 18 Pro family and, especially, Apple's first foldable iPhone will set the tone for his tenure.

It is a dual test. On the one hand, hardware: hinge, screen crease, and durability—ground where Samsung holds years of accumulated advantage and where Apple arrives, once again, late, but with the intention of doing it better. On the other hand, credibility: a new engineer CEO cannot afford a launch with mechanical problems.

The choice of a hardware engineer at the peak of AI software euphoria is, in itself, a statement. Apple is saying that its advantage will not lie in having the best model, but in having the best place to run it: 2.5 billion devices with proprietary silicon and data that does not leave the phone.

If that bet pays off, the conversation stops being 'who has the biggest model' and becomes 'who has the cheapest inference closest to the user.' If it fails, Apple will have paid for its delay with permanent dependence on Google, which is exactly the scenario Jobs would have considered unacceptable.