Apple barely talked about AI at its big iPhone 17 event

Apple’s hotly-anticipated Apple event was chock-full of news about updates to AirPods, the Apple Watch, the new suite of iPhones, and more. But during the one-hour-and-fifteen-minute presentation, which was considerably shorter than usual, there was one notable buzzword missing from a good chunk of the fanfare: AI.

Although CEO Tim Cook said during the livestream that the company was “taking the biggest leap ever for iPhone,” there were only cursory mentions of Apple Intelligence during the new iPhone unveiling.

Yes, the phones represent “advancements in Apple silicon, hardware, and software,” per Apple, meaning they’re better for gaming, photos, speed, battery life, and more. But most of the specific, consumer-facing AI tools Apple touted — like visual intelligence features and live translation in iMessage and FaceTime — were big talking points back in June during WWDC 2025. And they’re also not necessarily novel, in that Apple competitors, such as Google and Samsung, introduced comparable features a year ago or more.

The presentation was a far cry from Apple’s heavy-handed mentions of AI at the iPhone 16 unveiling, which led to public disappointment when some flagship features didn’t roll out as promised.

This year, Apple talked more about how AI is helping power features in the background, and less about how it’s putting AI in front of consumers’ faces — unlike Google’s Pixel 10 unveiling last month and Samsung’s event in January. Apple is too behind on its agentic AI efforts to make assistant tools front-and-center in the same way, so it focused a lot of this event on hardware and how AI helps power things behind the scenes.

Executives talked about how an updated neural engine powers Apple Intelligence, and how local large language models power better game play at higher frame rates. They mentioned how Apple was now building neural accelerators into each GPU core in order to provide “MacBook Pro-levels of compute in an iPhone,” making intense AI workloads possible.

But when mentioning how AI relates to its new AirPods, Apple leaned heavily on live translation and heart rate monitoring instead of taking Google’s approach with the new budget Pixel Buds 2A, where it discussed how they can be used to speak with Gemini AI. Executives mentioned how the devices’ advanced computational model combines with Apple Intelligence models running on iPhone to power live translation, and for heart rate sensors, they highlighted machine learning algorithms’ role in powering an on-device AI model for activity and calorie tracking, thanks to more than 50 million hours of training data from more than 250,000 participants in an Apple study.

The mentions of AI in the new Apple Watch presentation felt just as harried. Executives highlighted how Apple’s machine learning algorithms analyze, over 30-day periods, how users’ blood pressure responds to heartbeats. It’s thanks to a series of studies totaling more than 100,000 participants, and Dr. Sumbul Desai, Apple’s VP of health, said the company hopes to “notify over one million people with undiagnosed hypertension in the first year alone” and expects FDA clearance soon.

The AI arms race has never been more high-stakes for the companies vying for top spot — in part because they’ve invested so much in their AI efforts, and those efforts don’t come cheap. OpenAI was valued at $300 billion this year and reportedly expects to burn $115 billion through 2029. Anthropic recently raised $13 billion at a $183 billion post-money valuation. In the past few months alone, Meta has spent billions of dollars hiring top industry researchers after a more than $14 billion investment in Scale AI. And that’s just to name a few.
Apple has long been criticized for falling behind in the AI race. Part of that has been reflected in at least 10 reported exits from its AI research department recently, including four last week. Its lead for robotics research, Jian Zhang, reportedly departed for Meta, and three other AI researchers reportedly left Apple’s foundation models team, with two going to OpenAI and one to Anthropic.