Question 1

What is Apple Intelligence and how does it work?

Accepted Answer

Apple Intelligence is Apple's integrated AI system launched on October 28, 2024, initially in the US and later expanded to 200+ countries by May 2025. It operates on a hybrid architecture: a roughly 3 billion parameter on-device model runs directly on the iPhone's Neural Engine at 30 tokens per second with 0.6 millisecond latency, handling tasks like text summarization, notification prioritization, and Writing Tools. For more complex queries, requests are routed to Apple's Private Cloud Compute infrastructure, which uses custom Apple silicon servers with stateless computation, no logging, and no admin access. Apple Intelligence also integrates third-party models including OpenAI's ChatGPT (since December 2024) and Google's Gemini (since January 2026) for tasks that exceed on-device and PCC capabilities.

Question 2

Why is Siri still behind Google Assistant and ChatGPT?

Accepted Answer

Siri's major overhaul, which was originally expected in 2025, has been delayed to spring 2026. The delay stems from a combination of technical debt and leadership turnover. Apple's former AI chief John Giannandrea was replaced by Amar Subramanya, a hire from Google's Gemini team, in a move widely interpreted as an acknowledgment that Siri's existing architecture needed a fundamental rewrite rather than incremental improvement. With the new Gemini integration announced January 12, 2026, Siri's multi-turn conversational accuracy has improved to 87%, up from 52% under the previous system. Apple is essentially rebuilding Siri on top of a 1.2 trillion parameter custom Gemini model that runs on Apple's own Private Cloud Compute servers rather than Google Cloud, preserving the privacy architecture while gaining model capability.

Question 3

What is Apple Private Cloud Compute and why does it matter?

Accepted Answer

Private Cloud Compute (PCC) is Apple's cloud AI infrastructure built on custom Apple silicon servers. Unlike traditional cloud AI services from Google, Microsoft, or Amazon, PCC enforces stateless computation — meaning user data is processed but never stored, logged, or accessible to Apple employees. There is no remote admin access and no persistent storage of queries. Independent security researchers have verified the architecture. PCC matters because it allows Apple to run larger AI models (beyond what fits on-device) while maintaining the privacy guarantees that differentiate Apple from competitors. The Gemini deal announced in January 2026 runs on PCC infrastructure, not Google Cloud, meaning Google never sees user queries. This is a structural advantage no other company can currently replicate at Apple's scale.

Question 4

What is Apple's deal with Google Gemini and how much does it cost?

Accepted Answer

Apple announced a deal with Google on January 12, 2026, to integrate a custom 1.2 trillion parameter Gemini model into Apple Intelligence. The deal is worth approximately $1 billion per year, with a total value of up to $5 billion over the contract period. The critical detail is that the Gemini model runs on Apple's Private Cloud Compute infrastructure, not on Google Cloud. This means user queries processed through Gemini never touch Google's servers and Google has no access to the data. Apple also maintains its existing integration with OpenAI's ChatGPT, launched in December 2024, under a different arrangement where Apple pays nothing and OpenAI gains distribution to Apple's user base. The dual-model approach gives Apple access to frontier model capabilities from two competing providers without building its own frontier model from scratch.

Question 5

How many devices support Apple Intelligence and which ones are compatible?

Accepted Answer

As of January 2026, Apple has 2.5 billion active devices worldwide, an increase of 150 million year over year, with approximately 1.5 billion active iPhones. Apple Intelligence requires an iPhone 15 Pro or later (A17 Pro chip or newer), any M-series iPad or Mac, and iOS 18.1 or later. This means the majority of Apple's installed base does not yet support Apple Intelligence, which creates a multi-year upgrade cycle opportunity. iOS 18 adoption stands at 82% of compatible iPhones, slightly below the 10-year average of 83.2%, but the hardware requirement is the real bottleneck. Apple's Neural Engine in supported devices delivers 35-38 TOPS (trillion operations per second), which is necessary for on-device inference at 30 tokens per second.

Question 6

Is Apple Intelligence driving iPhone sales or hurting them?

Accepted Answer

iPhone revenue hit $85.3 billion in Apple's Q1 FY2026 (the holiday quarter ending December 2025), up 23% year over year — the best iPhone quarter in the company's history. Total quarterly revenue reached $143.8 billion, up 16% YoY. While Apple has not directly attributed the sales increase to Apple Intelligence, the timing aligns with the feature's expansion to 200+ countries and the integration of ChatGPT. However, Apple's stock has fallen approximately 25% from its all-time high, erasing roughly $900 billion in market cap, driven by investor skepticism about Apple's AI competitiveness and multiple class-action lawsuits related to alleged overpromising on AI features. The disconnect between record hardware revenue and declining stock price reflects Wall Street's uncertainty about whether Apple Intelligence is a genuine platform shift or a marketing rebrand of incremental features.