Apple Vision Pro Lead Takes Over Siri: Can It Reverse the AI Gap?

According to the latest reports, Apple carried out an in-depth restructuring of its AI team in 2025.

[1][2]. This technical expert, who once led the core algorithms of Google Search, chose to take a step back after seven years at Apple.

and reports directly to Craig Federighi, Senior Vice President of Software Engineering[1][2][3]. This change in reporting hierarchy is significant—AI is no longer an independently operating “silos” but is fully integrated into Apple’s software development system.

Apple simultaneously made the following key adjustments:

Adjustment Area	Specific Changes
Siri Team	Merged into the Vision Pro and Software teams, placed under unified jurisdiction of Federighi
Robotics Project	Transferred to the hardware division, separated from the AI team
Search and Knowledge Graph	Taken over by Eddy Cue
AI Research Functions	The foundational model and machine learning research teams have been marginalized

Amar Subramanya, a former Google executive, has been appointed as Vice President of AI, succeeding Giannandrea[1].

---

Apple’s position in the AI field has been described as “disastrous” by multiple media outlets. An analyst commented: “Their AI strategy is a disaster. Look at OpenAI, Microsoft, Meta—every other major tech company is pouring resources into the AI race, while Apple is far behind. It’s like an F1 race is happening right now, and Apple and Cook are just sitting on the sidelines drinking cappuccinos.”[2]

:

Company	Estimated AI Investment (2024-2025)	Key Achievements
Microsoft	Over $100 billion	Full integration of Copilot, deep partnership with OpenAI
Meta	Approximately $60 billion	Llama open-source ecosystem, Reality Lab transformation
Google	Approximately $75 billion	Leading multimodal Gemini, surging AI orders for cloud business
Apple	Approximately $20 billion	Delayed Apple Intelligence, limited Siri functionality

Apple’s dilemma in the AI field stems from three structural contradictions:

1. Conflict between Privacy Philosophy and Technical Route
   • Apple adheres to an on-device AI route where “user data never leaves the device”
   • There is an order-of-magnitude gap in the parameter carrying capacity of large models compared to cloud deployment
   • Technically, it is difficult to meet Apple’s internal strict standards[2]

2. Confidentiality Culture Contradicts AI Research Needs
   • Giannandrea continued Google’s style, encouraging researchers to publish their findings
   • However, Apple has always been closed and confidential, leading to fierce clashes between the two cultures
   • Researchers’ enthusiasm has been dampened, and team cohesion has declined[2]

3. Severe Underinvestment in Resources
   • Giannandrea once requested funds to purchase NVIDIA GPUs for computing power, but the final amount was cut in half
   • Apple is willing to spend $700 billion on share repurchases to maintain its stock price
   • Yet it remains extremely conservative in investing in AI infrastructure[2]

2025 has become the year of “mass exodus” for Apple’s AI team:

Time	Departing Executive	Destination
August 2025	Pang Ruoming, Head of AI Foundational Models	Meta
September 2025	Robbie Walker, Head of Siri and Search	(Unreported)
October 2025	Yang Ke, Head of AI-driven Web Search	Meta
December 2025	Alan Dye, Vice President of Human-Interaction Design	Meta
December 2025	John Giannandrea	Retirement

These departures mark that Apple’s AI team has been “dismantled and rebuilt”[2].

---

After Rockwell took over Siri, he immediately launched an evaluation of multiple technical routes, covering Apple’s in-house models as well as third-party solutions such as Claude, ChatGPT, and Google Gemini[2]. Apple has confirmed partnerships with the following external partners:

Partner	Cooperation Content	Positioning
OpenAI	Integration of ChatGPT into Siri and Apple Intelligence	Handle complex queries requiring extensive knowledge background[3]
Google	Gemini model and cloud technology	Build Apple’s next-generation foundational model, replacing the in-house development route[2][3]

Apple stated that it has not adjusted its agreement with OpenAI, but the introduction of Google Gemini will clearly reshape Apple’s AI technology stack[3].

Apple plans to launch a major upgrade in the iOS system update in early 2026:

• All-new Siri
  : Rebuilt based on large language models (LLMs), with the ability to call on-device foundational models
• App Intents
  : Execute complex multi-step tasks
• Private Cloud Computing Integration
  : Call external models (such as Gemini) to handle web search tasks when they cannot be completed on-device[1]

However, this timeline has been significantly delayed compared to the original plan. Apple Intelligence was originally scheduled to launch with iOS 18, but was only reluctantly rolled out with iOS 18.1; in March 2025, Siri’s new features were again postponed to 2026, with reports stating that Apple internally described the delay as “ugly” and “embarrassing”[2].

---

As the head of Vision Pro, Rockwell has the following traits that may help turn the tide:

1. Productization Experience
   : Successfully brought Vision Pro from concept to market, demonstrating the ability to integrate complex hardware and software
2. Cross-departmental Coordination
   : The Vision Pro project involves numerous technology stacks, cultivating Rockwell’s cross-domain leadership
3. User Experience Orientation
   : Vision Pro emphasizes immersive interaction, which is highly aligned with the goals of AI assistants (contextual understanding, proactive service)

However, Rockwell has inherited a hot potato:

Challenge Dimension	Specific Issues
Technical Debt	Siri’s underlying architecture is outdated; integration with modern LLMs requires a complete overhaul
Team Reconstruction	Severe loss of core researchers requires re-recruitment and integration
Competitive Window	OpenAI, Google, and Microsoft have already established significant first-mover advantages
User Expectations	The user disappointment accumulated over 15 years of Siri requires time to repair
Privacy Constraints	Apple’s privacy commitments limit the choice of technical routes

Analysts believe that Rockwell may adopt the following strategies:

1. Focus on Differentiated Scenarios
   : Leverage Apple’s hardware ecosystem advantages to seek breakthroughs in on-device AI experiences
2. Rapid Integration of External Capabilities
   : Use Google Gemini and OpenAI to narrow the gap with competitors
3. Redefine Product Value
   : Reposition Siri from a “voice assistant” to a “personal intelligent hub”

---

As of the close on January 14, 2026, Apple’s stock price was

, with a market capitalization of approximately [0].

Time Horizon	Performance
1 Day	-1.44%
5 Days	-0.69%
1 Month	-7.54%
3 Months	+3.84%
6 Months	+23.33%
1 Year	+10.29%

Apple’s stock price has fallen 5.06% year-to-date, underperforming the Nasdaq Index over the same period (down approximately 0.29% in the past 5 days)[0][4].

Core Metric	Value	Industry Comparison
Price-to-Earnings (P/E)	34.35x	Tech industry average is approximately 28x
Price-to-Book (P/B)	52.18x	Relatively high
Net Profit Margin	26.92%	Excellent
Operating Profit Margin	31.97%	Excellent
Return on Equity (ROE)	1.64%	Relatively low (due to large cash reserves)

Rating Category	Number	Percentage
Strong Buy	1	0.9%
Buy	67	61.5%
Hold	34	31.2%
Sell	7	6.4%

: The average target price is $312.50, representing a upside from the current stock price; the target range is $220-$350[0].

:

• January 13, 2026: Wedbush maintains “Outperform” rating
• January 12, 2026: Wedbush maintains “Outperform” rating
• December 17, 2025: Morgan Stanley maintains “Overweight” rating
• December 9, 2025: Citigroup maintains “Buy” rating

---

: After Rockwell takes over Siri, Apple is expected to leverage the capabilities of Google Gemini and OpenAI to launch a significantly enhanced Siri in 2026. However, due to R&D cycles and integration complexity, it will be difficult to reverse the gap with OpenAI and Google in the short term. It will take 2-3 years for Apple to restore its brand perception in the AI field.

: If Rockwell successfully integrates external AI capabilities and finds a differentiated path for on-device AI, Apple may establish a unique advantage in on-device AI experiences. Apple’s privacy protection commitment may become a differentiated selling point, attracting user groups sensitive to data security.

: Apple’s success or failure in the AI field will depend on the following key factors:

1. Whether it can achieve breakthroughs in in-house AI capabilities
2. Whether it can deeply integrate AI capabilities into core products such as the iPhone, iPad, and Vision Pro
3. Whether it can provide leading AI experiences while maintaining privacy commitments

Focus Area	Recommendation
Investment Rating	Maintain the “Buy” consensus, but be patient with AI progress
Risk Warning	AI progress falling short of expectations may become a valuation-suppressing factor
Key Watchpoint	Earnings conference call on January 29, 2026 (FY2026 Q1)
Core Logic	Apple’s hardware ecosystem + service revenue remain resilient; AI is only a long-term catalyst

Analysts generally believe that Apple’s AI dilemma will not have a substantive impact on its core businesses (iPhone, services) in the short term, but in the long run, if Apple cannot maintain competitiveness in the AI era, the differentiated advantages of its hardware products will gradually weaken.

---

[1] Sina Tech - “Report: Apple Restructures and Expands AI Team, Focuses on Launching All-New Siri Experience for iPhone 17 and Other Devices” (https://finance.sina.com.cn/tech/digi/2025-12-24/doc-inhcwafv4707878.shtml)

[2] Phoenix Net Tech - “Cook’s Retirement Imminent, Apple Enters the Eve of a ‘Nokia Moment’” (https://tech.ifeng.com/c/8puRVLENyl1)

[3] Yahoo Finance - “Siri Leadership Change? Apple’s AI Strategy U-turns to Partner with Google Gemini” (https://hk.finance.yahoo.com/news/siri背後換人-蘋果ai策略大轉彎找上google-gemini-173004947.html)

[4] Jinling AI - Market Data API [0]