product updateApple

Apple Releases Beta of Redesigned Siri in iOS 27 Ahead of iPhone 18 Launch

TL;DR

Apple released iOS 27 on Monday, introducing a beta version of its redesigned Siri AI assistant ahead of iPhone 18 availability this week. The new Siri uses large language models to search personal data like texts and calendars, but requires opt-in and isn't yet available in the EU or China.

3 min read
0

Apple released iOS 27 on Monday, bringing a beta version of its long-delayed, redesigned Siri AI assistant to iPhone users just days before the iPhone 18 lineup arrives in stores. The rollout is opt-in only, and Apple said some users may need to join a waitlist before gaining access.

The new Siri uses large language models to search a user's local personal data — including texts, emails, and calendar entries — to answer queries. According to Apple, the assistant can, for example, scan messages to surface addresses or recipe instructions, answer general knowledge questions, and search the web. A new standalone Siri app saves query history, and users can now reply to Siri's responses in a back-and-forth conversational format, a shift from earlier single-turn versions of the assistant.

Apple CEO John Ternus, who succeeded Tim Cook on September 1, called it "personal intelligence that's actually personal" during the company's iPhone launch event last week — his first keynote appearance as chief executive.

Smaller models, cloud offload, and Google infrastructure

Unlike frontier labs such as OpenAI, Anthropic, and Google, Apple relies primarily on smaller, in-house models designed to run directly on iPhone and Mac silicon. For queries too complex for on-device processing, Apple routes requests to its Private Cloud Compute system, which according to the company runs on Google Cloud infrastructure using Nvidia and Intel chips. Apple has also confirmed it is using Google's Gemini models to help build out its own AI systems — a notable dependency on a direct AI competitor's technology.

Apple said cloud-dependent Siri and Apple Intelligence features will be subject to daily usage limits that vary by "feature, request complexity, system demand, system policies, and other factors." The company said expanded limits will be offered for a fee at an unspecified future date — pricing not yet disclosed.

Limited language and regional availability

At launch, the redesigned Siri supports English only, with French, Japanese, Korean, Portuguese, and Spanish support planned for next month. Apple said the feature will not be available in the European Union or China initially, citing regulatory reasons — a familiar pattern for Apple Intelligence features, which have faced repeated regional rollout delays tied to the EU's Digital Markets Act and China's AI approval requirements.

The redesigned Siri was first announced in June, following a delay of the originally planned 2024 version. That earlier version, unveiled alongside Apple Intelligence, was pushed back after Apple said it needed more time to meet its own quality standards.

What this means

This is a beta rollout of a feature layered on Apple's existing model stack, not a new foundation model release — Apple isn't shipping new weights so much as shipping a new interface and orchestration layer on top of small on-device models, Private Cloud Compute, and third-party Gemini infrastructure. The real story is architectural: Apple is now openly reliant on Google Cloud and Gemini to deliver its flagship AI feature, undercutting the narrative of a fully vertically integrated Apple AI stack. The opt-in beta gate, waitlist, missing EU/China availability, and vague daily usage limits all signal a cautious, capacity-constrained launch rather than a confident, finished product — consistent with a feature that slipped by roughly two years from its original 2024 target.

Source: cnbc.com ↗

Related Articles

product update

Gemini app adds in-chat editing, full-screen preview and text highlighting for generated files

Google announced on October 29, 2026 that Gemini app users can generate, preview and edit documents, PDFs, presentations, spreadsheets, code and other files directly within a chat. The update adds in-place editing, full-screen preview and text highlighting to the file-generation capability Google introduced in April.

product update

Doubao Work adds Kimi K3.1 Pro, GLM-5.3 and DeepSeek-V4.1-Flash for team and enterprise plans

ByteDance's Doubao Work said on October 29 that it is opening third-party models to business customers. Kimi K3.1 Pro, GLM-5.3 and DeepSeek-V4.1-Flash are the first batch, available on team and enterprise plans and billed against existing AI credits.

product update

GitHub makes repository plugin defaults and enterprise plugin controls generally available in Copilot

Repository plugin defaults and enterprise-managed plugin controls are now generally available in VS Code, Copilot CLI, and the GitHub Copilot app, according to GitHub. Trusted repositories can install and enable default plugins.

product update

AGIBOT unveils Worken, a framework to measure robots' productive work after deployment

AGIBOT introduced Worken, a proposed framework for measuring how much productive work robots perform once deployed, at AGIBOT Innovation Day 2026 in Shanghai on October 29, 2026. The company also debuted the GO-3 Preview embodied foundation model and the WITA-Omni interaction model.

Comments

Loading...