Apple's AI Strategy Proves Cautious Approach Beats Racing Forward on Artificial Intelligence

Why Apple's slow-and-steady AI bet is starting to look pretty smart

For years, Apple has been accused of being one of the biggest stragglers in the AI arms race. Doubters have argued that Apple's lack of a clear AI strategy have cost it its edge, and Wall Street analysts have worried that the gap could start hurting iPhone sales. Now, the company has unveiled what it is billing as its biggest AI launch to date: Siri AI, which embeds new automated capabilities (fueled by a partnership with Google Gemini) into the very spine of its software. Is it enough to get people to stop saying that Apple is "losing" the AI race? To be honest, nobody really knows. But the question itself may be the wrong one. A better one might be: are Apple customers actually going to use these features and, if they do, will it help Apple's business? Before we address that question, we should note that Monday's announcements also came with an interesting comment from Craig Federighi, Apple's senior vice president of software engineering. "Some appear to be racing forward, seemingly pursuing AI for the sake of AI, without clear regard for the people -- all of us -- that it's ultimately meant to serve," Federighi said during his remarks. "At Apple, our mission has always been to turn the potential of advanced technology into helpful and intuitive products for everyone." The not-so-veiled defiance on display here seems like both a response to Apple's "behind-on-AI" criticism and an effort to acknowledge the deeply ambivalent -- and, according to some polls, increasingly negative -- sentiments that many consumers have about the AI industry. It's also a shrewd message at a moment when Americans are worried that AI will take their jobs and rot their brains. Apple is positioning itself as the AI company that's actually on your side. Judging by Monday's demos, that positioning has some substance behind it. Siri can now surface information buried deep in your inbox or text history and surface helpful information and offer helpful suggestions based on it. It can use what Apple calls onscreen awareness to give you context about what you're looking at. And -- using Gemini -- it can pull near-instantaneous up-to-date information from the web and deliver it right to your device. Siri is also designed to work seamlessly across Apple devices, giving users increased flexibility and, like other AI chatbots, it stores chat histories so users can revisit past conversations. By building AI functionalities into its disembodied, ethereal assistant, Apple also has the potential to eat into the advantages of competitors whose apps can only reach users through its own App Store. For those competitors, having Apple's AI embedded at the operating system level is a meaningful threat to their distribution advantage. The keyword here is "potential" since this version of Siri won't be available to consumers until later this year, as a beta. A final verdict will have to wait, but what's already clear is that Apple is doing its best to court its audience -- whether they end up going for it or not. Apple is obviously a hardware company, and these updates are designed to make that hardware incrementally more user-friendly and convenient, keeping users glued to their devices a little while longer. The contrast with its competitors is instructive and maybe the most important signal in Monday's announcements for anyone watching where the AI industry is actually headed. Take OpenAI, which, despite shipping updates at a relentless pace, has struggled to define who it's actually selling to, oscillating between consumers and enterprises. Or Meta, which is pouring gargantuan sums into AI without a clear explanation of how it connects to the company's core advertising business. Apple's more measured approach is starting to look optimal by comparison -- and more financially sound. For the most part, Apple hasn't needed a gangbusters AI strategy. It posted historic iPhone sales last quarter. And as questions mount over AI's profitability and real-world utility, Apple is spending significantly less than its competitors -- roughly $14 billion in capex planned this year, against a cumulative $900 billion being committed by other tech giants -- while still earning huge amounts of revenue. That revenue has come from the AI industry itself via taxes on AI companies that use its App Store to platform their apps. In short, Apple is spending less, making more, and now launched a suite of AI features that -- for many iPhone users -- will feel indistinguishable from the other AI applications already available to them through the App Store. If that doesn't exactly count as "winning the AI race," it may be the smartest way to run it.

Apple's Cautious AI Strategy Could Have Been Its Smartest Move

Macy has been working for CNET for coming on 2 years. Prior to CNET, Macy received a North Carolina College Media Association award in sports writing. You've probably heard someone say something along the lines of, "Apple is losing the AI race" or "Apple's AI is lagging." What if being behind on AI was intentional? Or maybe even a winning strategy? For years now, the media has treated Apple's caution in integrating AI as a failure, approaching every underwhelming developer event as a catastrophe. That framing failed to assess what the company was doing behind the scenes. Apple has spent decades evaluating how to bring innovative tech to real people. Apple didn't need to sprint into flashy demos that are loud on stage but don't appeal to everyday users. Choosing when to adopt a technology is a business strategy, not an act of cowardice. As tech consumers, we're also in an embarrassing contradiction. We've been asking Apple to hurry along with its AI integration, while also insisting we're exhausted by AI hype. We keep demanding novelty, and then complain about pointless novelty. During yesterday's WWDC keynote event, Apple's senior vice president of software engineering, Craig Federighi, gave the standard line about AI, calling it an "incredibly powerful technology with the potential to shape society in profound ways, and with proper care, unlock meaningful benefits for people everywhere." But he also implied that Apple is not about chasing trends that offer little value. "Some appear to be racing forward, seemingly pursuing AI for the sake of AI, without clear regard for the people -- all of us -- that it's ultimately meant to serve." Seems reasonable. I also thought we didn't want performative or useless AI from tech companies. I also thought we were becoming allergic to AI washing. Apple plays coy when it comes to AI The WWDC keynote rhythm and language mattered, too. Artificial intelligence barely showed up in the first half of the keynote, with the first mention coming right around the 28-minute mark of a 1-hour event. Only after the glossy parts had settled did Federighi stroll in and say that Apple will not race ahead with AI unless it's private and helpful. That line landed because it mapped onto Apple's major advantage: consumer trust. We're right to worry about how our data is being grabbed and sold to bad actors. We're right to worry about how AI can compromise our safety and security. Companies like OpenAI, xAI and Meta have lost public trust and had to make major policy changes to convince us that our sensitive information is being protected. Apple is deliberately careful with the term "AI." At least someone in the company must know that a lot of us freeze or check out when we hear it. Sparing vocabulary is a strategic decision aimed at avoiding alienation, and it's going to pay off for Apple in the long term. "WWDC 2026 is Apple's AI credibility test," Francisco Jeronimo, vice president of client devices at International Data Corporation, said in an email statement to CNET. "Apple does not need to win AI by having the biggest model or the loudest demo. It needs to make AI trusted, useful and invisible across the ecosystem." Apple isn't hemorrhaging money on AI Notice, too, that Apple treats AI models as commodities to be licensed or leveraged through partnerships rather than a cause for declaring war on competitors. There's no need to get behind every AI investment unless there's a real market for it -- and it proves to be profitable. That last point is crucial. The current AI hype, with its massive cloud compute, training and infrastructure requirements, is expensive. Shockingly so. A website called Is AI Profitable Yet? tracks how much frontier AI companies have invested versus how much they've made back in revenue. If the large red bars are any indication, big players like Amazon, Alphabet (Google) and Meta are not seeing a profitable return on their AI investments, costing them billions. (In the few minutes I spent on the page, $20 million had been spent on AI.) Meanwhile, Apple has avoided spending hundreds of billions of dollars on data centers and compute, which gives the giant some padding if -- or when -- the bubble pops. Apple's path looks a lot like this: Invest where there's a clear payoff and where privacy aligns for users -- and punt on the rest until it makes sense. Rather than going all-in on AI devices, products and services, Apple is betting on the growing use of local AI functions and on-device foundation models, particularly with the company's high-performance chips. If the AI bubble bursts, who survives? So, who survives the worst of the bubble? The companies that built consumer value rather than viral demos, leaned into privacy and trust as selling points and avoided massive, irreversible infrastructure bets. The survivors retained massive amounts of cash on promoting products that sell, and only made AI an ingredient -- not an identity. Apple seems well-positioned to survive an AI crash. If the market corrects all the cloud splurge and flashy demo mania, Apple will be standing with tools that work for people on their devices and a solid reputation (i.e., brand loyalty) that most of its rivals lack. That doesn't mean Apple is immune to mistakes or slow-to-fail products, but it does mean it's less at risk of overspending or misreading consumer demand. Apple still needs to prove that these on-device models and its revamped Siri provide value for us. But for now, the company is picking its battles and its timing.

As OpenAI leans into enterprise business, Apple and Google set sights on the masses

Apple and Google are developing AI features for their billions of users even though it's not clear if consumers are willing to pay for the enhancements. As OpenAI steers away from the consumer focus that made ChatGPT a household name, Google and Apple are rolling out a slew of new consumer AI offerings, trying to show how the technology can be practical for everyday users. The opposing approaches were laid bare this week, as Apple used its annual developers conference to introduce Siri AI as a new stand-alone app, and OpenAI announced that it's confidentially filed to go public, a move made possible by its recent traction in the enterprise market, largely in AI-assisted coding. The diverging tracks come at a pivotal moment in artificial intelligence, as OpenAI and Anthropic focus on building big -- and eventually profitable -- businesses by selling into enterprises that are eager to spend, rather than trying to lure paying consumers who are accustomed to free online services. Apple and Google, by contrast, have massive piles of cash and can afford to subsidize consumer use of AI if it means ramping up adoption and ensuring that coveted users stay in their ecosystem. Gartner analyst Kjell Carlsson said that for Apple, it's a matter of, "I can give this away for free, because I'll make it up on the iPhones or iCloud subscription they'll be buying." Apple says it has more than 2.5 billion active devices worldwide. Google currently has seven products that each serve more than 2 billion monthly users. "Companies are realizing users get value from AI through these products, experiences, and the solutions that we build with them, not necessarily through the models or platforms," Carlsson said. While Apple is finally showing some progress in consumer AI, its Worldwide Developers Conference was widely viewed as underwhelming given how late the iPhone maker is to the game and how much anticipation there's been for an upgraded Siri. The stock dropped more than 5% over two days, as analysts questioned the lack of concrete timing and delays in certain parts of the world. In addition to the new Siri app, Apple also showcased how AI is now being integrated in various products, such as the iPhone camera, email, and the Shortcuts automation and productivity app. The company also spent a good chunk of its keynote presentation on Monday showing new child safety tools, which are rapidly gaining importance as AI becomes ubiquitous. Apple's annual event came less than a month after Google I/O, the search company's high-profile developers conference. There, Google showed off a number of consumer AI products such as Gemini Spark, which is a general-purpose AI agent, and information agents that the company says operate in the background in search and "will send you an intelligent, synthesized update, with the ability to take action." Google also unveiled smart glasses, an effort to crack a corner of the wearables market where rival Meta has found success, and a video editing tool that lets users "change what's happening" in a clip they shoot. Google and Apple are longtime rivals in consumer technology, but they're also partnering in AI. Gemini is powering Apple Intelligence, the technology behind the new Siri. And Apple executives said at WWDC that Google and chipmaker Nvidia are helping the company with its most advanced model, called Apple Foundation Model Cloud Pro. Apple didn't provide a comment for this story. A Google spokesperson said that the consumer slant at I/O had to do with the nature of the event and that the company had hundreds of enterprise-focused announcements in April at its cloud conference. Almost all of OpenAI's announcements this year have been enterprise-driven, as the creator of ChatGPT and pioneer in generative AI now finds itself chasing Anthropic. Anthropic, which was founded by early OpenAI researchers, was valued at $965 billion in its latest funding round in May, topping OpenAI's $852 billion valuation from March. Anthropic also beat OpenAI to the confidential initial public offering filing phase, disclosing its move a week before OpenAI did the same. Last month, OpenAI announced the creation of OpenAI Deployment Co., or DeployCo, a joint venture majority owned and controlled by OpenAI alongside 19 global investment firms, consultancies and systems integrators. Its stated goal is to deploy "forward engineers" directly into corporations to bridge the gap between model capabilities and complex corporate workflows. OpenAI also agreed to acquire AI consulting and engineering firm Tomoro, which included 150 "deployment specialists." Meanwhile, OpenAI has abandoned some consumer products as it tries to rightsize its financials. In March, the company shuttered its video generation tool Sora, which hit 1 million downloads less than five days after its launch in late September. The same month, OpenAI announced a pivot away from the Instant Checkout shopping feature it launched last year. Denise Dresser, OpenAI's chief revenue officer, said last month that the company is at a "tipping point" in enterprise AI adoption, after CFO Sarah Friar said in March that enterprise was up to 40% of total revenue and would be at about half by the end of the year. "If you look at the total value of software, the vast majority of it is business software," said Rob Collie, founder of consulting firm P3 Adaptive and a former business intelligence lead at Microsoft. "That's where we make profit. That's where productivity is worth paying for." OpenAI used the popularity of ChatGPT to build its brand. But the real money is currently being spent in the AI coding market, where developers and non-technical people are using the company's Codex and Anthropic's Claude Code to write software and build apps based on text prompts. "Enterprise buying cycles are complicated and coding is the easiest funnel for companies to get into since engineering teams are blowing their budgets," said Ram Bala, associate professor of AI and analytics at Santa Clara University. OpenAI didn't provide a comment for this story. One particular risk that Apple and Google face in targeting consumers is that AI skepticism is running hot, due to fears that it's rapidly replacing jobs and leading to troubling behavior among children and teens. A Pew Research Center study published in March found that about half of Americans felt that AI in their daily lives made them "more concerned than excited." Alphabet CEO Sundar Pichai said in a recent episode of the Hard Fork podcast that people are "rightfully" anxious about what sort of future the technology will create, calling the scale of change unprecedented. Collie of P3 Adaptive said a "backlash" is happening, but that companies "perceived as friendly" could benefit from changing the narrative. With the entire tech industry almost singularly focused on AI and with Wall Street rewarding what it views as the AI winners and punishing the laggards, companies are investing as if the technology is inevitable and it's just a matter of who gets there first. "They've all learned the hard way the cost of missing a segment," Collie said. Gil Luria, a tech analyst at D.A. Davidson, said that even with OpenAI's race to capture the enterprise, the company still has a big lead over Google and others in the consumer market because of the viral success of ChatGPT. He said Apple's rollout of a Siri app "could very well attract a lot of consumers away from both ChatGPT and Gemini." And analysts at JPMorgan Chase wrote in a note on Tuesday that Apple's addition of expressive voices in Siri "could set up for a device upgrade cycle if these features gain strong consumer traction." Apple still has a lot to prove, and that task is soon to fall in the hands of incoming CEO John Ternus, the company's longtime hardware boss who's replacing Tim Cook at the helm in September. Matt Rogers, the co-founder of Nest and a former iPhone engineer under Steve Jobs, said Ternus has a high hill to climb. "Apple played it safe," said Rogers, who's now CEO of waste prevention company Mill, regarding the WWDC announcements. "As John Ternus takes over, he needs to steer the company towards making AI useful, trusted, and native across the devices people already live with." Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.