Apple software chief: Context-sensitive Siri was not a vaporware

The “context -sensitive siri” announced last year was only a video demonstration, i.e. Vaporware? Apple’s software chief Craig Federighi has now commented on the topic in several interviews on the WWDC 2025 and emphasized that it was definitely real software.

Confusion about architectures

The Hardware website Tom’s Guide the manager said Apple developed two versions of the new Siri, V1 and V2 internally. “We had completed the first version here shortly before the conference (WWDC 2024, editor’s note) and at the time we were very confident that we could deliver it.” It was planned to bring out the software until December, “and if not until spring”. Then they had been dealing with V1 for months and “better and better and better over more app intents and search processes”. “In principle, however, we found that the restrictions on the V1 architecture did not reach the quality level that our customers needed and expected.”

Since Apple did not seem possible to reach the Apple standards with V1, it was then decided to switch to the V2 architecture. “When we recognized this – that was in spring – we announced that we could not publish it yet and that we would continue to work on really switching to the new architecture and then bringing out something.” Apple does not communicate when the time has come: Federighi said at WWDC 2025 that something was reported next year.

Functions comparatively simple

Federighi did not say what the concrete difference between V1 and V2 was. However, it is clear that there were massive conversions internally: Apple exchanged the team management and Federighi made the project more directly. The new Siri actually deals with comparatively few new functions: the voice assistant should become more context-sensitive, i.e. access to data on the device, such as calendars or iMessage chats, which can then be used. Furthermore, Siri should have access to screen contents, which Apple now sometimes realizes with a screenshot function as part of Visual Intelligence in iOS 26 (also with the help of chatt). Finally, Apple had announced that Siri should also be able to interact with apps – via so -called app Inents, the developers should incorporate.

At Apple, Apple has not yet officially talked about a real chatbot or a voice assistant with communication skills, as you know about the language mode of chatt or google. An LLM-based Siri should only be released in 2026 or even 2027-if at all. There was no speech on the WWDC 2025. In another Interview on Siri with the Wall Street Journal Federighi tried to appease here: The topic of AI was a “long-term transformation wave” for industry and society that takes decades. “There is no reason to hurry with the wrong functions and the wrong product just to be the first.”