Apple is moving toward building its own AI infrastructure, but it is leaning on Google in the short term. The company plans to mass-produce its own AI-focused server chips in the second half of 2026. At the same time, it is using a partnership with Google to meet near-term expectations around Apple Intelligence and Siri.
Apple-Google Deal
According to analyst Ming-Chi Kuo, Apple faces immediate pressure in AI.
Kuo wrote in a post on X that Apple has two short-term problems. First, Apple needs a strong AI showing at WWDC after earlier promises about Apple Intelligence and Siri did not fully land. Second, cloud-based AI is improving so fast that user expectations have changed. Even if Apple delivers everything it previously announced, it may still fail to stand out without access to more powerful large-scale models.
As Kuo put it, “this year’s WWDC AI showcase can’t afford another disappointment,” and rising expectations now create “an urgent need for a more powerful on-device AI model.” Because of this, Apple is using Google’s models to close the gap.
A Temporary Strategy
Kuo said the Google partnership is not a long-term shift. It is a way to reduce pressure while Apple continues its own work. He also noted that on-device AI is unlikely to boost hardware sales right away. Still, he expects AI to become central to hardware design, the operating system, and the overall user experience.
To underline that point, Kuo compared Apple’s situation to Tesla. “If Tesla’s FSD algorithms came from other vendors, its competitive advantage and valuation would almost certainly be lower,” he said. Control over core AI matters.
Kuo added that Apple’s in-house AI server chips should enter mass production in the second half of 2026. Apple-run data centers are expected to follow in 2027. This timeline suggests that demand for Apple’s own on-device and hybrid AI will grow more meaningfully from 2027 onward.
For now, Apple is buying time. Long term, it wants to own the technology that defines its AI future.
