Apple has put a lot of effort into AI this year. Next year will be a crucial year

One of the biggest launches apple’s History is scheduled to be announced next year, but it has nothing to do with hardware.

The company promised investors it would launch the next generation of Siri, its artificially intelligent voice assistant. There’s a lot going on with Apple announcing a “more personal Siri.” Siri has so far stayed out of the tech industry’s AI race that began when OpenAI announced ChatGPT in late 2022.

Apple typically doesn’t release its product roadmaps to the public, but in the case of Siri, the iPhone maker made an exception. The company had planned to launch a new AI assistant in 2025. But in March, Apple delayed the upgrade to sometime “next year” even after advertising the feature.

Consumers are becoming increasingly accustomed to having free-flowing conversations with the likes of ChatGPT and Anthropic’s Claude. Google There is pressure to catch up with Apple when it comes to Gemini chatbots.

CEO Tim Cook told investors in October that Apple has made good progress with Siri, which has “meaningfully raised the bar for what to expect,” said Gene Munster of Deepwater Asset Management.

“They basically said, don’t worry about AI this year, and next year we’ll surprise you with what we’ll show you,” Munster said.

Apple’s stock is up 12% so far in 2025, but much of that rise has come in recent months as the company’s iPhone 17 impressed investors when it launched in September. But Android maker Google is at the center of the AI ​​boom with its proprietary models and tensor-processing AI chips, and its stock has soared more than 60% this year.

Throughout 2025, AI was present everywhere in Silicon Valley, with the exception of Cupertino.

OpenAI released Sora 2, a video generation app that briefly topped the Apple App Store charts. Anthropic has released several new Claude models. Amazon has revamped its Alexa AI assistant. microsoft In November, it released software that allows companies to manage “AI agents” (a term used to describe AI programs that can run for hours at a time). flat MetaThe company, which is facing a shift in its own AI strategy, has moved to prepare for the release of its next Frontier model, codenamed Avocado, CNBC reported last week.

And at the beginning of this year, Nvidia It robbed Apple of the title of most valuable technology company. The driving force was Nvidia’s insatiable demand for graphics processing units. That year, the chipmaker began shipping a type of AI computer called the Grace Blackwell NVL72, which combines 72 individual AI GPUs. The price is estimated at $3 million.

Meanwhile, Apple hasn’t made a major AI announcement since 2024, when it announced Apple Intelligence. This software suite includes an image generator, text rewriter, ability to summarize push notifications, and integration with ChatGPT.

But so far, consumer reaction to Apple Intelligence has been mixed.

While the company’s improved AI-powered notification filtering and photo editing features have been praised, other AI features have faced issues. For example, Apple temporarily turned off (and then turned back on by default) an AI feature that inaccurately rewrote push notifications from its News app.

The most notable Apple Intelligence feature was an upgrade to Siri, which the company said would take longer to develop than originally thought, and was postponed to the spring. Greg Jozwiak, Apple’s head of worldwide marketing, said in a June interview with the Wall Street Journal that the company “didn’t want to disappoint our customers.”

AI was barely mentioned at the company’s global developer conference in June.

Apple said its new chip has improved AI performance and debuted a number of machine learning features, including live translation and intelligent call screening for AirPods. Developers were also invited to take advantage of Apple’s foundational model. However, the company made no announcements about the scale of the chatbot and generative AI products that its peers are releasing.

To close out the year, Apple reorganized its AI leadership position, positioning the company to execute on its strategy to keep up with rivals’ breakthrough technologies.

Apple announced in early December that John Giannandrea, the company’s head of machine learning and AI strategy, will retire in 2026. Many of his responsibilities will be shared among chief operating officer Sabi Khan, head of services Eddie Cue and new employee Amar Subramanya, who previously worked at Google and Microsoft. Apple said head of software Craig Federighi will also have expanded oversight of AI, with Subramanya reporting to him.

Subramanya’s hire is particularly noteworthy. He was head of engineering at Google Gemini before briefly joining Microsoft as an AI executive. iPhone makers don’t tend to talk publicly about their engineering talent, especially new hires who don’t report directly to Mr. Cook.

Announcements of vice president hires like Subramanya show how important it is for Apple to prove to investors and the public that it is willing to shake up its AI leadership.

It has become clear that Apple is taking a different strategy from its peers, which are taking a cloud-based approach to AI that requires significant infrastructure costs.

Google, Microsoft, Meta, and Amazon have collectively committed $380 billion in capital spending this year, much of it in Nvidia-based data centers to create and serve cutting-edge AI models.

Apple also increased its capital spending, but on a much smaller scale. Apple spent $12.71 billion on capital expenditures in the year that ended in September, an increase of 35% on an annual basis but less than what it spent in 2018. The iPhone maker has said that rather than using Nvidia chips for servers for Apple Intelligence, it is using chips originally designed for its own computers for user privacy reasons.

A key question for Apple is whether it will find a partner to power its new Siri.

Improvements to the upgraded Siri will include features such as the voice assistant making intelligent reservations based on a user’s travel plans and personal relationships.

Now, when Siri is presented with a complex query, the AI ​​suggests ChatGPT to answer the question. In a panel discussion shortly after Apple Intelligence’s launch last year, company executives said other underlying models, including Google’s Gemini, could be incorporated into the service. The latest version of Google’s model, Gemini 3, was released in November to positive reviews.

Cook also said Apple is open to making major acquisitions, something that has been extremely rare in the past. The valuations of AI labs like OpenAI and Anthropic have reached levels that make acquisitions nearly impossible, even for companies with the cash flow of Apple.

OpenAI reached a valuation of $500 billion with a stock sale in October, and Anthropic reached a valuation of $350 billion in November. By comparison, Apple’s largest acquisition in history was its $3 billion purchase of Beats Electronics in 2014.

Apple’s lack of spending has some investors concerned about the company’s AI strategy and increased pressure for Siri upgrades.

“Investors already have enough gray hairs to wait for Apple to come up with an AI strategy,” said Wedbush analyst Dan Ives. “It’s time to go out and show the world what the strategy is.”

While some investors worry that Apple is falling behind in AI, the company’s most important business is doing better than ever.

The iPhone 17 has been a hit, and the company expects sales to increase 10% in the holiday quarter. According to Counterpoint Research, Apple is likely to surpass Samsung in shipments in 2025 and next year to become the top smartphone vendor.

Counterpoint analyst Yang Wang said Apple’s lackluster AI has not yet had a negative impact on iPhone sales, adding that other tech companies’ new AI features have not yet dramatically changed the daily experience of using smartphones.

“We don’t think it’s a big threat to Apple yet, just because competitors haven’t completely crushed Apple,” Wang said.

Analysts and consumers may not recognize the threat to Apple, but company executives do. Apple’s Mr. Qiu testified at the trial in May, saying that advances in AI technology may mean users won’t need an iPhone within 10 years.

Because new hardware devices can use AI to create new user interfaces and features that aren’t possible on smartphones. Some early AI gadgets are already on the market.

Ray-Ban Meta glasses can use AI to identify objects within a user’s field of vision, and Meta announced this month that it had acquired a startup called Limitless. The company’s AI pendant can record conversations and generate summaries of them. The purchase price was not disclosed.

But Apple’s biggest threat may come from its current AI partners.

Earlier this year, OpenAI acquired former Apple design guru Jony Ive’s AI device startup io for $6.4 billion, and Ive is now helping the AI ​​Lab build the next generation of consumer devices. Ive, who left Apple in 2019, is still widely recognized as one of the driving forces behind the hardware maker’s biggest hits, including the iPhone and iPad.

OpenAI CEO Sam Altman said in November that the company had “finally” completed its first device prototype. Neither he nor Ive would say what the device is, just that it’s targeted more toward the gentle “vibes” of hardware than smartphones.

Earlier this month, Altman told journalists that he believes OpenAI’s real competitor is Apple, not Google. According to the Wall Street Journal, he said smartphones are not well-suited for AI companions and other use cases.

But Apple still has time to prepare its counters. Ive said in November that it would take about two years for OpenAI devices to become publicly available.

“They have more time than people realize to figure this out,” Munster said. “But in the short term, when this new Siri comes out, it should be a 10 out of 10.”