Apple Intelligence Gets a Comprehensive Upgrade as OS Advances to Version 27

Author | Yezi

Click to Listen to the Podcast Version of This Article

In the early hours of June 9, Beijing time, Apple CEO Tim Cook took the stage at the Worldwide Developers Conference (WWDC) for the last time as leader. This event, laden with a sense of "farewell" from the outside world, did not conclude with sentimentality or surprises. Instead, Apple delivered a nearly pragmatic software update, submitting a long-overdue answer in the generative AI race two years late.

From iOS 27 to macOS 27, from the completely revamped Siri AI to the system-wide integration of Apple Intelligence, Apple used the entire event to declare: it is no longer absent from the AI era.

However, the capital market's feedback was a decline in stock price during the session, ultimately closing down approximately 1.9%. This complex market sentiment reflects Apple's true position in the AI competition: not redefining the industry, but engaging in a systematic catch-up.

Siri AI Finally Makes Its Debut: From Voice Assistant to System-Level Agent

The most core announcement at this WWDC was undoubtedly the complete revamp of Siri. Apple officially renamed it "Siri AI." It is no longer a pop-up voice assistant but debuts as a standalone app, with an interactive interface fully aligned with mainstream AI chatbots.

The revamped Siri AI boasts three core capabilities: cross-app operations, screen awareness, and personal context understanding. Live demonstrations showed that users could directly instruct Siri to find specific photos taken during a trip and share them with contacts without opening the Photos app. After taking a picture of a restaurant bill, Siri could directly identify the amount and split the bill evenly, all without switching between multiple apps.

On iPhone models equipped with the Dynamic Island, Siri's wake-up animation is integrated into the Dynamic Island area, and the chat interface can be activated directly by swiping down from the middle of the screen.

What shook the industry even more was the choice at the technical architecture level. Craig Federighi, Apple's senior vice president of software engineering, announced at the event that Apple has entered into a deep collaboration with Google, introducing Google's Gemini large model technology at the foundation of the new Apple Intelligence. This marks the first time Apple has leveraged an external company for its core AI capabilities, exposing its shortcomings in independently building ultra-large-scale cloud model capabilities.

Apple emphasized that, even with the introduction of external models, its privacy-first principle remains unchanged: the vast majority of computations are completed on the device, while complex tasks run through Apple's Private Cloud Compute framework, with data used solely for processing user requests and not stored or used for third-party model training. According to media reports, Apple pays approximately $1 billion annually for this collaboration, though the company has not confirmed this.

This decision breaks with Apple's narrative tradition of trying to reduce its reliance on Google over the past decade. Since removing Google Maps and launching its own mapping service with iOS 6 in 2012, Apple has consistently emphasized full-stack self-research. However, in the AI era, faced with the rapid progress of OpenAI and Google in large models, Apple has chosen a more pragmatic strategy, leveraging external model capabilities while safeguarding system access points, privacy boundaries, and user experience.

AI Capabilities Fully Integrated, But Mostly Catching Up on Shortcomings

Beyond Siri's revamp, Apple has also embedded AI capabilities into its operating system. The Photos app introduces three major professional AI tools: Extend (Smart Expansion) automatically pushes photo boundaries outward; Reframe (Spatial Reconstruction) allows users to change photo composition by holding and dragging the subject; and the CleanUp (Removal) function has been significantly rebuilt, resolving previous issues with incomplete object removal. Additionally, a new natural language photo editing capability has been added, allowing users to modify specific parts of images through voice commands.

Safari browser, through Apple Intelligence, can automatically organize tabs by content type and frequency of use and monitor webpage updates, actively notifying users when page content changes. The Shortcuts app has also been reshaped by AI, allowing users to describe their needs in natural language, with the system automatically generating corresponding shortcuts without the need for manual module assembly.

The Camera app introduces a "Siri Mode" for quickly invoking Apple Visual Intelligence capabilities, such as identifying plants, translating text, reading nutritional labels, and capturing contact information from business cards. On the Mac side, Visual Intelligence allows users to circle content on the screen for direct text and image recognition and search.

A noteworthy detail is that Apple has also, for the first time, allowed users to set third-party AI services like ChatGPT and Claude as their default response engines, achieving so-called "model freedom." This open stance further confirms Apple's strategic shift in the AI race, no longer attempting to control everything but using its ecological integration capabilities as its core competitiveness.

However, market analysis generally believes that most of the functions announced in this release were actually promised two years ago when Apple Intelligence was first unveiled, equivalent to completing "catch-up" work rather than introducing disruptive products beyond market expectations. Compared to the established leading positions of OpenAI and Google, Apple is still in a catching-up position.

iOS 27: A "Snow Leopard"-Style Pragmatic Evolution

The core development philosophy of iOS 27 (internal codename Rave) is defined by Apple as "Snow Leopard"-style, focusing not on piling up flashy new features but on optimizing full-chain system performance and enhancing underlying stability. Apple claims to have streamlined redundant code, reducing memory usage by 20% and extending daily battery life by 1 to 2 hours. App launch speeds on iPhone and iPad have increased by up to 30%, photo display speeds in the library have improved by 70%, and AirDrop speeds have increased by 80%.

In terms of compatibility, iOS 27 supports all device models running iOS 26, making it the iOS major version update with the broadest device coverage to date. However, the full Apple Intelligence and Siri AI are only supported on iPhone 15 Pro series and later models, iPads, and Macs equipped with M1 or higher chips.

The most significant change in macOS 27 "Golden Gate" is that the system will cease support for Intel-based Mac computers, only compatible with devices equipped with M1 or newer Apple-designed chips. This marks the conclusion of the six-year Apple Silicon transition, with Intel Macs officially becoming a thing of the past.

Child Safety and Parental Controls: Strengthening the Family Ecosystem's Business Logic

Another noteworthy change at this WWDC is that Apple devoted considerable time to introducing children's and family safety strategies. iOS 27 and macOS 27 introduce new children's accounts and enhanced parental control features.

Parents can easily create children's accounts, with the App Store defaulting to a confirmation mechanism for purchases and browsing for users under 13. The system has upgraded time quota management, supporting fine-grained control over app usage permissions by time period. iMessage and Safari have added real-time detection of explicit content.

Apple has also introduced new APIs, requiring all apps to adapt to age-appropriate privacy protection experiences. New time limits and official recommendations for gaming, social, and other apps have been added, allowing parents to flexibly adjust "time quotas" and ensure their children stay focused during school hours by setting app availability schedules.

This is not just a routine security update. From a business perspective, it helps Apple strengthen its family ecosystem—when a family's life is bound to the Apple ecosystem, the cost of switching to another phone brand becomes higher. Apple is not just trying to lock in a single user but an entire family.

Chinese Mainland Market: AI Features Temporarily Unavailable

For Chinese users, this event brought disappointing news.

Federighi explicitly stated that in the Chinese mainland, due to the need to comply with regulatory requirements and advance relevant compliance work, Apple Intelligence will not be available for now. The Siri AI beta test version will debut exclusively in the U.S. market this fall, with the EU also temporarily excluded due to unresolved regulatory plans.

This means that domestic Apple users will be unable to experience the core AI functions announced in this release for a considerable time. This has also created an awkward reality on Chinese internet platforms where this year's WWDC is seen as "upgrading AI abroad while upgrading UI in China."

This regional "AI ecosystem temperature difference" poses a direct challenge to Apple's commercial prospects in China. As domestic large models like Qwen and DeepSeek accelerate their penetration into the smartphone bottom layer (Note: " bottom layer " is kept as is for now due to potential context loss in translation; clarify if a specific term is needed) and move toward Agent implementation, the iPhone's appeal in the high-end market, lacking core AI capabilities, faces the risk of erosion.

Market Reaction Lukewarm: Apple's AI Still Faces Three Major Challenges

The capital market reacted indifferently to this AI answer. Before the event, Apple's stock price once (yídù, meaning "once" or "at one point") rose nearly 3.3%, but as the announcements progressed, gains gradually narrowed and turned negative, ultimately closing down approximately 1.9%. The stock price continued to decline in after-hours trading.

Analysis suggests that the market's lukewarm response stems from three aspects: first, some functions are still in beta, with the new Siri not opening for testing until this fall, still far from large-scale commercialization;

second, compared to competitors, the level of innovation is limited, with many functions, such as AI assistants, visual recognition, or writing tools, already available in products like Google's Gemini and ChatGPT;

third, global expansion is hindered by regulatory restrictions, making it difficult for Apple's AI ecosystem to achieve synchronous global expansion in the short term.

For Cook, this event marks his last WWDC in an 11-year tenure as CEO. As planned, Cook will officially step down on September 1, succeeded by John Ternus, senior vice president of hardware engineering.

In his closing remarks, Cook stated, "Looking back on my years as CEO, events like this are among the most brilliant moments. Seeing what you've created with the tools we provide reminds me time and again that imagination knows no bounds."

Industry views point out that the significance of WWDC 2026 for Apple lies in finally completing the foundational framework of its AI ecosystem, fully integrating AI capabilities into its hardware and software systems.

However, over the next year, the actual user experience of Siri AI, the global rollout speed of Apple Intelligence, and the progress in developer ecosystem construction will ultimately determine the success or failure of Apple's AI catch-up. Apple may not need to own the strongest model, but it cannot allow others' AI to become the first access point on iPhones and all its hardware products. This is perhaps the most core strategic proposition Cook has left for his successor.

THE END

Copyright and Disclaimer

1. Content Copyright: Except for quoted public data, policies, and cases, this article is original. Professional data is sourced from authorized databases and government websites, with cases compiled from real events.

2. Image Licensing: Some images in this article are original or officially licensed, while others are AI-generated; for network images without clear copyright, credit belongs to the original authors, and infringement claims will be removed upon notification.

3. Republication Guidelines: Unauthorized republication is prohibited; republication must retain the complete source and author.

4. Liability Disclaimer: This article is a commercial figure observation and industry commentary compiled by the author based on public information, for reference only, and does not constitute professional advice. Risks arising from use are borne by the user. Xinzhidongneng reserves the final interpretation rights to this article.