Former Apple chief executive John Sculley says the company’s next leader should pivot from apps to agentic artificial intelligence, a signal of how fast AI strategy is changing in Big Tech. His comments raise questions about Apple’s long-term direction after years of services growth, and how it can challenge OpenAI and other AI leaders.
Sculley, who led Apple from 1983 to 1993, argued that a future CEO after Tim Cook will need to rethink priorities. He suggested a move away from App Store-centric thinking and toward systems that can act on a user’s behalf across devices and services.
The Call for a Strategic Shift
“CEO Tim Cook’s successor should focus less on apps, and more on agentic AI to compete with OpenAI.”
Sculley’s view reflects a wider debate about the next phase of consumer technology. Apps helped drive Apple’s services revenue for over a decade. But agentic AI—software that plans, decides, and executes tasks—could redraw how people get things done on phones and computers.
OpenAI and others have been testing agent-like tools that can read emails, book travel, and use third-party services with minimal input. Apple has highlighted privacy-preserving on-device processing with its Apple Intelligence initiative announced in 2024, including a smarter Siri and optional ChatGPT access for some tasks. Yet Sculley’s point suggests Apple may need to go further with full, goal-directed assistants.
Background: From Apps to Assistants
The App Store transformed Apple’s business after 2008, spawning a developer economy and new subscription services. Under Cook, Apple expanded services, wearables, and high-end hardware, while leaning into privacy as a brand pillar. AI features grew incrementally, often focused on photography, dictation, and suggestions.
By 2024, Apple formalized a broader AI approach. Apple Intelligence combined on-device and cloud models and introduced writing tools, image features, and a more context-aware Siri. It also integrated ChatGPT for certain requests with user consent. The company framed this as practical AI that respects privacy.
Agentic AI raises tougher questions. To act for users, systems need wider permissions, deep context, and reliable links to calendars, payments, and messaging. That can clash with privacy limits and platform policies designed to protect users from abuse and errors.
What Agentic AI Could Change
Agentic systems aim to complete goals, not just answer questions. That could reduce reliance on tapping through many apps. It might change the economics of mobile software if value shifts from standalone apps to agent skills and trusted workflows.
- Users may prefer one assistant that handles bookings, shopping, and support.
- Developers might build agent actions rather than full apps.
- App discovery could give way to task discovery inside assistants.
For Apple, a deep agent model would need tight integration with Siri, Messages, Mail, Maps, Wallet, and third-party services. It would also require strong safeguards, audit trails, and clear user control to avoid mistakes and misuse.
Competition and Risks
OpenAI has invested in models that plan and use tools. Google and Microsoft are pushing assistants that work across search, productivity, and cloud. Apple’s advantage lies in device integration, custom silicon, and privacy features. But rivals move quickly in cloud AI and developer ecosystems for agent workflows.
There are risks. Agent errors can create costs, fraud exposure, or safety issues. Clear consent and transparency are needed when an assistant spends money or sends messages. Regulators are watching how assistants access data and whether platforms unfairly favor their own services.
Analysts say Apple must balance privacy with capability. Too much restriction could limit agent usefulness. Too little could erode trust. Striking the right balance will determine whether users adopt agent features at scale.
What to Watch
Key signals in the months ahead include deeper Siri upgrades, expanded partnerships for third-party actions, and new developer tools for agent skills. Any shift in App Store policy to support agent workflows would be notable. Hardware announcements that boost on-device reasoning and memory could also hint at a stronger agent focus.
Sculley’s call highlights a turning point. Apps defined the smartphone era. The next phase may be about assistants that can plan and act. Apple’s path—shaped by privacy, design, and ecosystem control—will show whether it can match, or even surpass, the agent push from OpenAI and others.
