Apple's AI Strategy Comes Into Focus at WWDC
Apple's developer conference next week will reveal how the company plans to fix its consumer AI strategy-and whether a decade-old voice assistant can power its comeback in the category.
The Worldwide Developers Conference begins June 8 with Apple facing a different challenge than usual. Instead of showcasing new software features, the company must demonstrate a working AI strategy after Apple Intelligence, launched in 2024, fell short of expectations. The experience was incomplete and underwhelming, despite Apple's promise of privacy-first AI across its product ecosystem.
Apple's new CEO, John Ternus, who takes over in September, inherits a company that arrived late to consumer AI and has since fallen behind on foundational models and infrastructure. His task: pivot the company toward a compelling alternative to competitors without abandoning Apple's core strengths.
Siri Gets a Second Chance
Apple is betting on Siri, its voice assistant, to anchor the AI comeback. The modernized version, built on Google's Gemini AI technology, will remember user queries and access data across devices for personalized responses.
Siri has become a running joke in tech-more useful for setting timers than acting as a true assistant. Apple's redesign matters because Siri is the missing piece for future devices: smart glasses, camera-equipped AirPods, wearables, and the Apple Watch. Without a working AI assistant, Apple struggles to compete in the next wave of AI hardware.
How Siri integrates with Apple's first-party and third-party apps on iPhone and other devices-and how developers can tap into it-will be worth watching at the conference.
Outsourcing Intelligence, Owning Integration
Apple chose not to build its own large foundational models. Instead, it partnered with Google to power Siri and other AI features. This marks a departure from Apple's traditional ecosystem lock-in strategy.
The move is not necessarily a strategic failure. Francisco Jeronimo, Vice President of Client Devices at IDC, said Apple is not behind on AI features available to users-it's behind on frontier models and infrastructure. "Outsourcing the model layer is not necessarily a bad strategy, provided Apple owns the integration and the privacy architecture on top of it," he said.
Apple has made similar choices before. It did not build its own browser or maps app. Instead, it focused on providing the best user experience through integration. The same logic applies to AI: Apple wants to control distribution and user experience, not necessarily build the underlying infrastructure.
The Hardware-AI Gap
Apple has invested in AI-powered silicon for years, believing heavy workloads should run on-device using chips inside the iPhone or Mac. But AI is evolving toward architectures that rely on cloud servers for seamless functionality.
This creates a design challenge. Apple must build AI systems that leverage its hardware to enable functional AI without constant cloud connectivity. The company also needs to extend beyond screen-based devices into screenless ones like smart glasses, where AI relies on camera awareness and natural voice interaction.
Anshel Sag, Principal Analyst at Moor Insights & Strategy, outlined the formula: "Make AI easy to use; don't make it any more invasive than it needs to be. If Apple can integrate its AI really well into iOS and create a connected fabric across devices, it can win."
Current Apple Intelligence remains limited to basic tasks. Neither Apple nor the industry has figured out how average consumers actually want to interact with AI.
Apps or Agents?
The industry is moving toward AI agents-systems that use other software applications on behalf of users, handling spreadsheets, calendars, and email without explicit commands. Apple has not clarified its position on agents.
The traditional app store model, where users browse, install, and learn interfaces, feels outdated in an era where large language models function as the operating system's brain. Apple may need an AI agent store to stay relevant.
The bigger question: will consumers actually want personal AI agents, and will they solve problems that current apps do not? Customer service remains one of the largest pain points users face. Whether AI agents and chatbots can address it remains unclear.
Privacy Versus Personalization
Apple built its brand on privacy-first user data handling-a differentiator that allowed it to avoid targeted advertising models competitors rely on. How this stance evolves under Ternus remains uncertain.
The tension is real. Advanced personalization requires understanding user data. Apple must decide whether to maintain its privacy-first approach or embrace the data collection necessary for truly personalized AI.
The Design Problem
Industry insiders say Apple's core challenge is designing AI that makes judgments for users rather than forcing users to ask questions. The company understands consumers and creates seamless experiences. Now it must apply those skills to AI.
AI is already useful for technical users-software developers use Anthropic's Claude to write code. On the consumer side, AI still lacks clear use cases that regular people want. This is where Apple must outperform every tech company: designing AI that works for everyday problems without requiring users to learn new workflows.
What Apple shows at WWDC will signal whether the company can solve this problem or whether the gap between AI capability and consumer need remains too wide.
For product developers: Understanding AI for Product Development and how companies like Apple approach Generative AI and LLM integration provides insight into the real constraints and opportunities shaping AI product strategy in 2024.
Your membership also unlocks:
