A YEAR ago, when Apple used a jamboree at its home in Silicon Valley to unveil its artificial-intelligence (AI) strategy, grandly known as Apple Intelligence, it was a banner occasion.
The following day the firm’s value soared by more than US$200 billion – one of the biggest single-day leaps of any company in American history. The excitement was fuelled by hopes that generative AI would enable Apple to transform the iPhone into a digital assistant – in effect, Siri with a brain – helping to resuscitate flagging phone sales. Twelve months later, that excitement has turned into almost existential dread.
It is not just that many of last year’s promises have turned out to be vapourware. Siri’s overhaul has been indefinitely postponed, and Apple Intelligence is no match for other voice-activated AI assistants, such as Google’s Gemini. Meanwhile, Apple’s vulnerabilities in China have been exposed by US President Donald Trump’s trade war. Moreover, it faces new legal and regulatory challenges to the two biggest parts of its high-margin services business.
Its shares, down by almost a fifth this year, have lagged behind its big-tech peers, Alphabet, Amazon, Meta and Microsoft. But those are not the most alarming comparisons.
In a new book, Apple in China, Patrick McGee draws an ominous parallel between Tim Cook, Apple’s chief executive, and Jack Welch, boss of General Electric from 1981 to 2001. Like Welch, Cook has made a fortune for investors – when Apple’s market value first exceeded US$3 trillion, in 2022, it had risen by an average of more than US$700 million per day since he took over from Steve Jobs in 2011. But McGee raises the possibility that, as at GE, Apple’s success may obscure serious vulnerabilities. If that is the case, what can Cook do to avoid the sort of fate that befell GE, Nokia and other great companies that suddenly lost their way?
The answer is unlikely to emerge during Apple’s annual Worldwide Developers Conference that starts on Jun 9. Amid reports of upheaval among executives, it is expected to return to its unflashy roots of announcing software updates for its phones and computers, rather than revealing a refreshed approach to AI.
BT in your inbox
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
Many would prefer to see Cook work on a new hardware strategy instead. Craig Moffett of MoffettNathanson, an equity research firm, notes that the greatest moments in Apple’s history have come from the reinvention of what techies call “form factors”: the Mac reimagined desktop computing, the iPod transformed personal music habits and the iPhone popularised touchscreen smartphones. AI looks like it will be another such pivot point. (Eddy Cue, Apple’s head of services, recently admitted that AI could make the iPhone irrelevant in ten years.)
For now, Apple’s rivals have been faster to explore new opportunities. Meta and Google are pinning hopes on AI-infused smart glasses, as are Chinese tech firms such as Xiaomi and Baidu. OpenAI, maker of ChatGPT, recently announced a US$6.4 billion deal to buy a firm created by Jony Ive, Apple’s former chief designer, to build an AI device. As yet there is only hype to go on, but it has put Apple’s lack of AI innovation in the spotlight.
Apple’s response may seem like dogged incrementalism. Next year it is expected to unveil a foldable phone, following a path blazed previously by the likes of Samsung and Motorola. But Richard Windsor of Radio Free Mobile, a tech research firm, thinks Apple may still have an ace up its sleeve. If smart glasses take off, its investment in the Vision Pro virtual-reality headset, though so far an expensive flop, may be an insurance policy. It could provide Apple with enough expertise in headgear and eyewear to shift quickly to glasses. If so, the company will avoid “doing a Nokia”, he says.
Likewise, Apple might make use of this moment of soul-searching to rethink other shibboleths of Cook’s tenure, such as the obsession with privacy and the high walls it puts around its family of products. As Ben Thompson of Stratechery, a newsletter, points out, sanctifying the privacy of its users’ data has been an easy virtue for Apple to uphold because until recently it did not have much of an advertising business. Yet in the AI era, it has drawbacks.
First, Apple’s reluctance to scrape customers’ individual information makes it harder to train personalised AI models. Apple uses what it calls “differential privacy” based on aggregate insights, rather than the rich, granular data hoovered up by firms such as Google.
Second, privacy has encouraged it to prioritise AI that runs on its own devices, rather than investing in cloud infrastructure. Chatbots have advanced more rapidly in the cloud because the models can be bigger (awkwardly, this led Apple to offer some users of Apple Intelligence an opt-in to ChatGPT).
In order to overcome its AI deficiencies, it could splash out on buying a builder of cloud-based large language models (LLMs). But it has left it quite late.
OpenAI’s deal with Ive makes it less likely to ally with Apple. Anthropic is close to Amazon, which has a big stake in the maker of the Claude family of LLMs. Other options are either Chinese or too small for a company of Apple’s heft.
Alternatively, it could relax its “walled garden” ethos of seamless integration, and partner with a variety of third-party LLMs, as Motorola, owned by the Chinese firm Lenovo, has done. Third-party voice-activated chatbots could quickly solve its Siri problem, giving renewed reason for people to upgrade their phones.
The likelihood is that Cook will do nothing radical. As Moffett puts it, his tenure has been marked by the steady ascendancy of “process over product”. Instead of flashy innovations, his hallmark has been metronomic reliability, especially with regard to financial performance. Nor has he any hope of swiftly extricating Apple from China. As McGee points out, even if Apple’s final assembly moves to India and elsewhere, the supply chain’s roots remain deeply embedded in the Middle Kingdom.
Yet this is no time for complacency. Whatever the ups and downs of AI – as Google has recently shown, yesterday’s losers can quickly become today’s winners – nothing turns investors off quicker than a profits shock. That is what makes the threats to Apple’s services business so serious.
The most striking risk is that the judge who declared Google a monopolist may order it to suspend payments to Apple that make Google’s search engine the default on the iPhone. The payments, which are partly for exclusivity and partly a revenue-sharing arrangement, generate about US$20 billion a year for Apple (last year its services revenue was US$96 billion).
David Vogt of UBS Investment Bank says that, if the judge imposes a ban on the exclusivity part of the payments, it could cut Google’s revenues by about US$10 billion. “I’m getting calls every day of, ‘What will the market do to Apple stock if that happens?’” he says. Google has vowed to appeal.
Another looming threat is to app-store revenues, which are under scrutiny as a result of the European Union’s Digital Markets Act, as well as from an antitrust lawsuit brought by Epic, a gaming firm, against Apple in America. Bank of America estimates that app-store commissions generate US$31 billion a year of high-margin services revenue for Apple.
If app developers steer customers away from Apple’s app store as a result of the rulings, it could clobber the lucrative cash cow.
Services have been the brightest spot of Cook’s tenure in recent years, helping to mitigate stagnation in iPhone sales. It will certainly be a blow if the line of business suffers. But if it prompts Cook to tear up his own rule book on AI and everything else, it may be worth it in the end.
©2025 The Economist Newspaper Limited. All rights reserved
