A new rumor suggests Apple will introduce a web search feature backed by Apple Foundation Models that can call out to Google Gemini to enhance Siri's ability to gather and summarize information.
The rumors of Apple Intelligence's death have long been exaggerated, and a new rumor points to Apple's commitment to its internal models. While third-party models from Google and others will be involved via Private Cloud Compute, it seems even search features will be handled by Apple's models at the base.
According to a report from Bloomberg, Apple is planning to release its app intent-powered Siri in the spring, and one of its features will be enhanced AI web search. It will be powered by Apple's Foundation Models that will run the contextual app intent system too.
That isn't to say third parties won't be involved, especially since Apple's search deal with Google will continue. Instead, Google has allegedly provided a version of Gemini that runs in Private Cloud Compute and helps act as the web summarizer tool.
The report suggests that the new Siri will have three core components: a planner, a search operator, and a summarizer. Apple's Foundation Model will act as the planner and search since that's dealing with on-device personal data, but getting the data from the web and collating it may be up to the Google model.
There is still a lot to learn about Apple's approach to AI going forward, as it had to scrap its previous approach entirely. The new Siri powered by Apple Intelligence LLMs is expected to launch in early 2026 with iOS 26.4.
Apple's vertical integration continues
The report is a sharp turn from previous ones, from the same publication, that suggested Apple could be planning to abandon its internal models. Instead, it seems to suggest what AppleInsider has been discussing all along — a cooperative effort between on-device Apple models and third-party models running via Private Cloud Compute.
It's not dissimilar from how Google search works with iPhone and Siri today. Google is the default search engine in Safari, and that can be changed by the user, but it's also the search engine used by Siri.
Some queries rely on something called Siri intelligence, which is an old term that predates AI and refers to algorithms derived from device and web data. For example, Siri and Spotlight can surface information about celebrities from the web or show info derived from a Calendar event.
When Siri hits a wall and can't produce an answer on its own, it conducts a web search via Google. However, user privacy is preserved since Siri is using a unique identifier so the query isn't attached to the user, their device, or their IP.
The upgrade to Apple Intelligence seems to take a similar approach. We'll have to see exactly how Apple lays it out, but from the recent reports, it seems on-device Apple Foundation Models will be responsible for parsing app intent systems and personal data.
These on-device systems will power contextual actions, system-wide suggestions, and more. However, when Siri and its new LLM backend detect that a user query will require additional resources, it'll call out to an AI agent equipped to deal with the topic.
Initially, it seems that system will be Google Gemini via a model running in Private Cloud Compute servers controlled by Apple. The so-called "world knowledge" will come from it and be summarized and presented to the user.
This system is different from how it works with Apple's ChatGPT partnership today. When Apple Intelligence passes a query to ChatGPT, it is run on OpenAI servers, though there is a contract in place that forces OpenAI to discard queries and data.
Slow and steady wins the AI race
Evidence is mounting that Apple is taking a particularly unique approach to artificial intelligence. Just like it didn't build a search engine to compete with Google, it's not going to build copycat bots that Google, Anthropic, and OpenAI can provide to users already.
Apple dominates markets it wasn't first in, and AI may be next
To top it off, Apple will maintain its promise of privacy, security, and leaving the planet better than we found it by keeping everything running on Apple technology. Private Cloud Compute servers are powered by renewable energy and are verifiable as not training against or storing user data.
There's no telling when Apple might discuss these new upgrades, but it may not be in 2025. The company has already had a PR hit from announcing the delayed contextual AI systems that were meant for the spring.
For now, we'll have to continue delving into the data provided by leakers and anonymous sources "familiar with the matter." If everything that has been rumored comes to pass, ignoring the doom and gloom angles around it, Apple will have quite a competitive offering in the AI space.
However, the general public is cooling on AI and the bubble may soon pop. If so, that may actually be a boon for Apple, as it hasn't built its entire business on flashy promises, but instead continues to focus on what is actually useful to customers.
As usual, Apple is often never first, but it eventually become best in class. Even Apple CEO Tim Cook believes its slow and measured approach to AI will make it become the best.
