Author: Hu Shixin; Editor: Ye Jinyan

Produced by DeepWeb & Tencent News Xiaoman Studio

Early morning Beijing time on June 9th, Apple's WWDC 2026 kicked off in Cupertino. Tim Cook, as usual, took the stage with a "Good morning," but this time it felt more like a farewell: this was his 15th and final time hosting WWDC as CEO. On September 1st, John Tenus, Senior Vice President of Hardware Engineering, will take over at Apple.

This year marks Apple's 50th anniversary, and its market capitalization has reached a high of $4 trillion; however, this 78-minute keynote address featured no new hardware products, focusing almost entirely on AI. Over the past two years, Apple Intelligence has been repeatedly delayed, the Siri redesign has yet to materialize, and Apple even paid a $250 million settlement in North America due to AI-related controversies.

ChatGPT has rewritten the industry for two and a half years. Can Siri, which was once disliked by Apple users, become the entry point for the iPhone again?

Apple's answer was to redesign Siri, leverage Google's model technology to build underlying capabilities, and embed AI into native apps like Safari, Photos, Passwords, and Shortcuts. However, the capital market didn't seem to buy it immediately. On the day of the launch event, Apple's stock price closed down 1.89%, and continued to decline by about 1.92% in after-hours trading, resulting in a market capitalization loss of over $75 billion.

Apple's AI, which the outside world has been waiting for for two years, has finally come to the forefront. The question is, has it truly addressed its shortcomings, or is it simply repackaging a belated arrival as a fresh start?

Siri AI, lending its "soul" to Google

About 30 minutes into the presentation, Apple Intelligence and Siri AI took the stage.

It is understood that the next generation of Apple Intelligence will be advanced through a collaboration with Google, building upon the technology behind the Gemini series models to create the next generation of Apple foundation models—Apple Foundation Models. Part of this model will run on the device, handling lower-latency tasks more relevant to personal data; the other part will be handled by private cloud servers, dealing with heavier requests such as image generation and complex reasoning.

Apple has also introduced a system orchestrator to schedule capabilities such as contextual understanding, world knowledge, app interaction, and screen awareness. In other words, Apple wants AI to do more than just answer questions; it wants to understand what the user is looking at, what's on the device, and which apps can be accessed.

Apple also reiterated its consistent privacy narrative: data only serves the current request, is not stored, and is inaccessible to Apple and third parties.

Following the press conference, Apple executives added that this is not simply a matter of integrating Gemini. Apple's base model consists of multiple models on the device and in the cloud, customized for Apple Silicon. While it borrows Gemini's distillation technology during training, the model ultimately running for users is Apple's own.

The redesigned Siri, now renamed Siri AI, has its own standalone app for the first time, and conversation logs can be synced across iCloud. It's integrated into the "Dynamic Island" feature, supports screen content understanding, and can use App Actions to invoke applications to complete tasks: such as generating party menus based on race schedules, retrieving information from text messages to add invitations, identifying and splitting bills, and determining whether a backpack is allowed on board based on flight information. Native apps like Safari, Password, Phone, and Photos also embed AI capabilities.

Apple's intention is clear: Siri is no longer just a voice entry point, but an operational hub embedded between the system and applications.

However, the feedback after the launch event wasn't entirely optimistic. The main point of contention was that while Siri AI did address its shortcomings in contextual understanding, screen awareness, and cross-application functionality, most of its capabilities were in areas that had already been repeatedly demonstrated by large-scale products over the past two years. It made Siri more like a system-level assistant than before, but it hasn't yet shown a truly unexpected new entry point.

Furthermore, due to regulatory requirements, Siri AI and the full Apple Intelligence functionality are not yet available in the EU and mainland China. For users in China, the most crucial AI update announced at the launch event will only be visible and not immediately usable.

Cook's AI debt

Apple's current passive position in AI is the result of gradual accumulation over the past decade or so.

Since Tim Cook took over Apple, he has almost brought the company to its commercial peak. Over the past 15 years, Apple's market value has grown from approximately $350 billion to $4 trillion. However, on the other hand, the deterministic management style that the "Cook era" excelled at has also made Apple slow in the generative AI race.

Senior analyst Ming-Chi Kuo previously commented that Cook built a profit wall with extreme supply chain management, but also burdened Apple with heavy AI debt.

Siri is the most typical entry point for this debt. In 2010, Steve Jobs acquired Siri for $200 million. In 2011, the iPhone 4S was released, and Siri debuted with it. It was originally Apple's earliest bet on a smart assistant, but it repeatedly missed upgrade windows in the following decade. The head of Siri changed several times, from Scott Foster and Eddie Cue to Federighi and John Giannandrea. Each adjustment was hoped to be a reboot, but in the end, it never made Siri truly smarter.

In 2018, Cook poached Giannandrea from Google to bolster Apple's AI capabilities. At the time, Apple's Senior Vice President of Software Engineering, Federighi, told the team that this was exactly the kind of AI talent Apple needed. But seven years later, Siri remains stuck in the awkward position of being "wakeable and able to answer questions, but not user-friendly." The emergence of ChatGPT truly brought the problem to the surface. An Apple executive later admitted to Bloomberg that before that, Apple's AI wasn't even a clear solution.

Apple attempted to salvage the situation. At WWDC 2024, Apple Intelligence was unveiled, showcasing Siri's ability to read personal data and operate the phone across apps. The iPhone 16 was marketed as "born for Apple Intelligence." However, after the new phone went on sale, the core version of Siri failed to launch as promised. Bloomberg later revealed that Federighi discovered during internal testing that some demo functions were unstable, and those impressive visuals were largely pre-recorded prototype videos. After repeated delays, Apple removed related advertisements, and iPhone 16 buyers filed lawsuits.

Apple subsequently restructured its AI team. Reportedly, in March 2025, Siri was separated from Giannandrea and handed over to Mike Rockwell, head of Vision Pro. Rockwell joined the core Vision Pro team, replacing some of the original Siri staff, and, along with Federighi and others, pushed forward with a collaboration with Google to strengthen the underlying capabilities of the new Siri using Gemini and Google Cloud. The Information revealed that the key reason behind the collaboration was that Apple's self-developed model was not yet ready, especially struggling to run stably on devices.

Organizational upheaval ensued. Key members of the foundational model team, including Pang Ruoming, left Meta. In late 2025, Giannandrea announced his retirement, and former Google executive Amar Subramaniam succeeded him as Vice President of AI, reporting to Federighi. Apple's traditionally stable product system appeared rushed for the first time in the face of AI.

This isn't the only missed opportunity during Cook's era. DeepWeb previously summarized several of his regrets: the "Titan" car project, which took ten years and cost nearly $10 billion, was ultimately cancelled; the Vision Pro boasted impressive technology, but its $3,499 price tag and limited application scenarios led to a lukewarm market response, with only about 390,000 units shipped in 2024. In comparison, Siri has caused even more anxiety.

Bloomberg reports that after Apple's AI debacle, Tim Cook has unusually become deeply involved in the AI ​​roadmap, participating in key decisions. At an all-hands meeting in August 2025, he stated that this was "somewhat an opportunity for us," and Apple would invest unlimited resources in it. One senior executive believes that Apple's past strategy of leveraging its billions of users to overtake competitors may not work this time.

(AI-generated image)

We can't create the absolute best model, but we also can't lose our entry point.

According to many analysts, what Apple most wants to retain in AI this time is not the "strongest model" label, but the primary entry point on the iPhone.

A developer who has long followed the Apple ecosystem believes that Apple is willing to accept model capabilities from external partners and to share some of the computing costs with model vendors, but it cannot accept users bypassing Siri and system services to directly use ChatGPT, Gemini, or Claude as the default entry point on their phones. What Apple wants to protect is the layer of system permissions that allow users to make requests, access data, and complete tasks.

This is why Apple Intelligence repeatedly emphasizes on-device computing, private cloud computing, and personal context. Compared to simple question-and-answer, Apple is more concerned with whether AI can understand the screen the user is looking at, the information already on the device, and which apps can be accessed. The aforementioned source believes that while third-party models can become increasingly powerful, they struggle to naturally access the deepest personal context within a user's phone—a part Apple is unwilling to reveal.

Private cloud computing is therefore not just a technical solution, but also related to commercialization. Federighi mentioned that some functions that rely on server models will have daily usage limits, with iCloud+ users receiving higher limits. Many analysts therefore predict that Apple may tie high-cost AI features to a subscription system in the future, but the premise remains that users are convinced that their data is not handed over to external modeling companies and that Apple has not sacrificed its privacy commitments.

The battle for market entry could also change the rules of the App Store. Bank of America analyst Wamsey Mohan suggests that if Siri evolves into an AI agent, future app competition may no longer focus solely on downloads and usage time, but rather on who can become the service Siri invokes. Mohan estimates that if Siri successfully transforms into an AI agent, it could bring Apple an additional $15 billion to $30 billion in revenue by fiscal year 2030.

The concerns of the cautious are also clear. MoffettNathanson analyst Craig Moffett believes the market has already priced in Apple Intelligence as a catalyst, not a potential risk. Apple's current valuation is higher than its five-year average; to support this valuation, it must prove that AI can drive a larger wave of device upgrades or translate into new service revenue. The capital market isn't looking for Siri to become better, but rather whether AI can revitalize Apple's growth trajectory.

Hardware entry points were originally seen as another hidden avenue for Apple's AI development. Multiple reports previously indicated that Apple had been pushing forward a series of wearable devices centered around Siri and visual perception, including smart glasses, pendant devices, and AirPods with cameras. These all pointed in the same direction: if AI entry points shift from the screen to vision, voice, and spatial perception, Apple cannot rely solely on the iPhone.

However, this path has not been smooth. According to multiple reports, Apple has adjusted some of its Vision hardware plans, shifting resources to lighter smart glasses; the AirPods project with a camera, although once thought to be close to late-stage testing, was subsequently reported to have been temporarily shelved due to EU privacy compliance and supply chain adjustments.

An industry insider who has long followed Apple believes that what Cook left for the next CEO is a system logic that needs to be rewritten in the AI ​​era. In the past, Apple could wait for the technology to mature and then create a better experience through the integration of hardware and software; but the AI ​​window does not completely follow this rhythm. It requires continuous iteration, user habits, and faster organizational response.

Siri's original co-founder, Dag Kitlaus, remains optimistic. He told Bloomberg that Apple still has a chance to make Siri the user's first choice again with a "brain transplant." This statement sounds simple, but it points to a core problem Apple has faced for the past decade: Siri has never lacked an entry point, but rather, it lacks sufficient intelligence.