Two years ago, Apple announced its first foray into artificial intelligence. Built largely on internally developed models, “Apple Intelligence” promised to transform the Siri assistant, previously considered limited, into an insightful personal assistant as capable as the most advanced chatbots, with the added benefit of access to the user’s personal data and other powerful features. The effort ended in an embarrassing failure, with the company delivering little of what it had promised.
The iPhone maker doesn’t need to build models to profit from the technology
Apple’s recently revealed new approach to Siri AI highlights a different strategy. Instead of relying exclusively on internal development of large-scale models, the company explores partnerships and integrations that allow it to monetize the technology more efficiently, taking advantage of its installed base of hundreds of millions of devices.
This update transforms Siri into a more conversational and integrated assistant, capable of understanding personal context, analyzing what is on the user’s screen and searching for updated information on the web. The focus on privacy, with on-device or private cloud processing, reinforces Apple’s unique position in the market.
Analysts point out that, by not needing to compete directly in building giant foundation models, Apple can focus on user experience and ecosystem, areas where it already dominates. This places it as an unexpected competitor, or “dark horse”, in the AI race, especially against rivals that invest billions in model training.
Availability begins with developer testing on iOS 27, iPadOS 27 and other systems, with beta for users later in 2026. The launch marks a second chance for Apple to deliver on the promises it previously made.
