Is AI Agent Apple's big opportunity?
In the next stage of AI investment, Apple may transform from a "bystander" to a core beneficiary—but this depends on Siri truly evolving.
According to Chase Wind Trading Desk, BofA Securities analyst Wamsi Mohan raised Apple’s target price from $330 to $380 in a research report dated May 26, maintaining a Buy rating. The core logic is: In the era of agentic AI, the value will flow to platforms that control user intent, personal context, app access, permissions, identity authentication, payment, and trust—not necessarily the party with the most powerful models. The iPhone happens to sit at the intersection of these elements.
This target price corresponds to 37x the expected annual earnings per share of $10.29 in 2027, previously 32x. The valuation upgrade is not based on short-term sales forecasts, but on Apple potentially regaining pricing power in the service layer, transaction layer, and entry layer in the agentic AI era. BofA estimates Agentic Siri could bring $15-30 billion in net incremental revenue by fiscal 2030, with bull scenarios reaching $40-65 billion and the EPS increment possibly approaching $2.
The biggest uncertainty follows: Siri has disappointed the market before. If users’ task habits shift to external assistants like ChatGPT, Gemini, or Claude, Apple will still control the hardware and permissions, but the AI value it can capture will be greatly reduced.
Apple’s AI opportunity is not “the strongest model,” but “the most trustworthy entry point”
The core experience of the first generation of generative AI is Q&A, summarization, writing, and code generation: users directly ask the model, and the model side is more likely to capture value. Agentic AI follows a different logic: it needs to plan steps, invoke tools, read context, request confirmation, and complete actions. A strong model alone is not enough; it also requires comprehensive coordination with calendar, contacts, email, messages, location, payment, identity authentication, and app permissions.
Smartphones are naturally at the intersection of these elements. Compared to emerging AI hardware, phones already have scale, an app ecosystem, payment rails, biometric authentication, and daily usage habits. Apple’s global smartphone shipments in 2024 were 232.1 million, with a market share of 18.7%; in Q4 2025, shipments will be 83.4 million, with the share rising to 24.8%. This is an entry point that requires no zero-education for users.
Agentic AI scenarios directly rewrite value distribution. Users no longer manually open calendar, email, maps, and wallet, but say, "help me change the meeting, notify attendees, update the route, and pay." Whoever controls the routing, permissions, and confirmations after this sentence has a chance to redefine the way search, app distribution, transactions, and ads are monetized.
The iPhone’s moat is chip plus iOS working together
Apple’s hardware advantage in agentic AI has two layers.
The first layer is Apple Silicon. Agentic tasks require frequent processing of voice, images, text, notifications, search, and intent classification. If all requests are sent to the cloud, latency, privacy, reliability, and costs become obstacles. Local inference shortens response times and reduces cloud inference expenses. The iPhone 17 series memory is upgraded from previous generations: iPhone 17 has 8GB, iPhone Air/iPhone 17 Pro/Pro Max have 12GB. To support a fully-enabled local AI experience, the report judges at least 16GB—and more likely 24GB—is needed; current hardware does not meet this level.
The second layer is iOS. Chips determine how much intelligence can run locally; iOS determines whether this intelligence can truly handle tasks for users. Agentic AI needs to access personal data, invoke apps, confirm user identity, execute payments—these are OS-level capabilities. Apple doesn’t need to own every model, but it can decide which tasks are handled locally, which go to Private Cloud Compute, and which are assigned to external models—as long as Siri is the default entry point, underlying models can be routed, replaced, or combined.
According to third-party sources, Siri may be heavily rebuilt in iOS 27, shifting from a command-driven voice assistant to a more persistent, conversational agent that can invoke personal context and act across apps. If true, this directly addresses Apple’s most needed weakness: not making a Siri that “can answer,” but a Siri that “can execute.”
Siri must become the execution layer of iPhone or the valuation story won’t stand
Five key pathways, missing any one calls the whole logic into question.
- The new iPhone must be a better AI device. Needs stronger NPU, higher memory, greater bandwidth, better cooling, and battery efficiency.
- Hybrid inference must work in economic terms. High-frequency, low-complexity tasks are handled locally; complex but privacy-sensitive tasks go to Private Cloud Compute; only some tasks sent to external models.
- Siri must understand intent, retrieve context, call apps, and complete workflows. This isn’t just an upgrade in voice recognition—it’s a system-level workflow overhaul.
- App Intents must become a broad framework for app actions. Third-party apps are not just opened by user clicks but invoked by Siri as needed.
- The trust layer cannot be lost. Which data can be read, which actions can be performed, which steps must be confirmed, which transactions need Face ID or Apple Pay verification—all must be controlled by iOS with fine granularity.
Real risks are concentrated here: Apple can control devices, OS, App Store, Apple ID, Face ID, Apple Pay, and privacy architecture, but if Siri’s execution ability is insufficient, users won’t entrust real tasks to it.
Today Apple drives growth through hardware, but may improve service monetization via “intent execution”
Apple’s current business model is centered on a hardware-driven installed base flywheel. In fiscal 2025, product revenue will reach $307 billion, about 74% of total; service revenue at $109.2 billion, about 26%. The iPhone remains the key entry point: FY2025 revenue $209.6 billion, higher than FY2024’s $201.2 billion.
The profit structure better explains the strategic motivation. In FY2025, product gross margin is 36.8%, service gross margin as high as 75.4%. Service delivers about $82.3 billion gross profit, 42.2% of Apple’s total, though only about 26.2% of income. This means any entry that increases service usage and transaction frequency is highly leveraged for Apple’s profit structure.
The strategic value of agentic AI is precisely this: it could push service monetization from “app access and transaction” to “intent routing and trusted execution.” Users may not open a specific app, but Siri can distribute requests to an app, merchant, payment tool, or model service. Apple holds the front door and gains a chance to participate in subsequent distribution, transactions, payments, ads, and subscriptions through the whole chain.
Six potential incremental revenue sources include: Apple Intelligence Pro subscriptions; fees for default model placement or routing; App Store agentic e-commerce commissions; Apple Pay/Wallet routing fees; ads in agentic results; and iCloud+ AI layer upgrades. After deducting 20-30% overlap between sources, the net incremental revenue base scenario is $15-30 billion, bull scenario $40-65 billion.
Cost issues will decide whether AI is profit accretive or a margin pressure
Agentic AI consumes far more tokens than traditional Q&A—task execution requires reading context, planning steps, invoking tools, confirmation, executing transactions, maintaining state, all exceeding simple Q&A in token usage.
Estimating high-usage scenarios: by 2030, iPhone AI users will be about 1.729 billion; each makes roughly 20 requests per day, each request about 5,000 tokens, annual cloud token consumption around 316 quadrillion tokens. Converted to infrastructure demand: about 1 billion tokens processed per second, needing 8-18 million M5 Max-level chips, total facility power about 0.7-2.5GW, potential capital expenditure $6.7-50.4 billion.
The range is highly elastic. Apple’s FY2025 data center and hosting electricity consumption is about 2.585 billion kWh, averaging about 295MW, which provides a foundation. The critical variable is task allocation ratio: if many requests use external models, AI acts more like a variable cost business—the more it’s used, the higher token spending, and the more service margins are pressured. If high-frequency tasks are mainly local and within Private Cloud Compute, plus subscriptions, upgrades, transaction commissions, and payment revenue, AI could then become a source of long-term profit growth.
The next step for App Store: from “downloading apps” to “invoking actions”
Agentic AI will also reshape the App Store’s role. Currently, the App Store operates around downloads, subscriptions, in-app purchases, search ads, and developer distribution fees. In the future, users might not search for an app but directly tell Siri “book a hotel,” “buy a ticket,” “file a reimbursement,” or “create a document”—the travel app supplies flight inventory, e-commerce platforms provide merchants/logistics, banks offer accounts/payments, productivity software supplies documents/workflows.
Competition logic then changes: from “which app gets downloaded” to “which app Siri invokes.” Apple’s monetization points shift accordingly—covering agentic app discovery, App Intent certification, transaction commissions, agentic behavior governance, developer analytics tools, and possibly a future model/agent market.
Developer willingness to cooperate is key. If developers see Siri integration as a new commission layer, adoption may be slow; if App Intents bring incremental demand and higher conversion rates, a new distribution flywheel may form. For Apple, lowering or waiving fees early and proving value through real transaction data may be more crucial than setting commissions from the outset.
~~~~~~~~~~~~~~~~~~~~~~~~
The above excellent content is from Chase Wind Trading Desk.
For more detailed interpretation, including real-time analysis and frontline research, please join [Chase Wind Trading Desk ▪ Annual Membership]
Risk Warning and DisclaimerThe market is risky; investment should be cautious. This article does not constitute personal investment advice, nor does it consider the particular investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Invest accordingly at your own risk.