Apple to Let Users Pick AI Models in iOS 27

Apple is set to transform its devices into open AI platforms by allowing users to select from a range of third-party artificial intelligence models across its software features. The sweeping change, expected to arrive with iOS 27, iPadOS 27, and macOS 27 this fall, marks a dramatic strategic shift for a company long known for its tightly controlled ecosystem.

Rather than locking users into a single AI provider, Apple will offer choices among multiple external models for tasks including text generation, image creation, and editing — positioning itself as a neutral orchestrator in the rapidly evolving AI landscape.

Key Takeaways

• Multi-model choice: Users will be able to select different AI models for different tasks directly within Apple's native software features
• Fall 2025 rollout: The changes will ship with iOS 27, iPadOS 27, and macOS 27
• Google partnership: Apple has partnered with Google to improve the foundational models powering Siri
• Anthropic collaboration: Apple is working with Anthropic to bolster its internal AI infrastructure and product development
• Platform strategy: Apple is pivoting from AI provider to AI platform, aggregating third-party capabilities
• User empowerment: The move gives consumers unprecedented control over which AI powers their Apple experience

Apple Bets Big on an AI Marketplace Model

Apple's decision to open its software features to multiple AI providers represents one of the most significant strategic pivots in the company's recent history. Unlike its traditional approach — where Apple designs, builds, and controls nearly every layer of the user experience — this new model embraces external AI capabilities as first-class citizens.

The implications are profound. Instead of competing head-to-head with OpenAI, Google, Anthropic, and Meta in a costly AI arms race, Apple is positioning itself as the premier distribution channel for AI services. This mirrors the strategy that made the App Store a $90 billion annual ecosystem — Apple provides the platform, and third parties compete to deliver the best experience.

For users, this means granular control over their AI experience. Someone might prefer Google's Gemini for research-heavy text generation, Anthropic's Claude for nuanced writing assistance, and another model entirely for image editing. iOS 27 appears designed to make these choices seamless and integrated, rather than requiring users to jump between standalone apps.

Google and Anthropic Land Key Apple Partnerships

Two partnerships stand out in Apple's emerging AI strategy. The company has struck a deal with Google to improve the underlying models that power Siri, Apple's voice assistant that has long lagged behind competitors like Google Assistant and Amazon Alexa in terms of intelligence and capability.

This Google collaboration suggests Apple recognizes that building world-class large language models from scratch requires resources and expertise that even a $3 trillion company cannot easily replicate. Google's Gemini family of models, which have shown strong performance on industry benchmarks, could give Siri the intelligence boost it desperately needs.

Separately, Apple has partnered with Anthropic — the maker of Claude — to support internal AI infrastructure and product development. This partnership appears more foundational, potentially helping Apple build better AI tooling for its engineers, improve code generation workflows, and accelerate the development of AI-native features across its product lineup.

The dual partnership approach is notable:

• Google handles the consumer-facing Siri improvements
• Anthropic powers behind-the-scenes infrastructure and development
• Apple retains control of the overall user experience and data privacy framework
• Neither partner gets exclusive access to Apple's 1.5 billion active devices

Why Apple Chose to Be a Platform, Not a Provider

Apple's pivot to an AI platform model did not happen in a vacuum. The company's initial AI efforts, branded under Apple Intelligence and launched with iOS 18, received mixed reviews. Critics noted that Apple's on-device models, while privacy-friendly, often lagged behind cloud-based competitors in raw capability.

The AI landscape has also shifted dramatically over the past 18 months. The gap between leading models like GPT-4o, Claude 3.5 Sonnet, and Gemini 1.5 Pro continues to narrow, making it increasingly difficult for any single provider to maintain a decisive advantage. By offering multiple models, Apple sidesteps the risk of betting on the wrong horse.

There is also a competitive dimension. Samsung and other Android manufacturers have already integrated Google Gemini deeply into their devices. Microsoft has embedded Copilot across Windows and Office. Apple needed a differentiated approach — and giving users the power to choose may prove more compelling than any single AI integration.

From a business perspective, the platform model opens new revenue streams. Apple could take a commission on premium AI subscriptions accessed through its devices, similar to its App Store revenue model. If a user subscribes to Claude Pro or Gemini Advanced through an iOS 27 integration, Apple would likely capture a share of that recurring revenue.

What This Means for Developers and AI Companies

For AI companies, access to Apple's massive install base represents an enormous opportunity. Apple's ecosystem includes approximately 1.5 billion active devices worldwide, and even a small percentage of users opting into a third-party AI model could translate to millions of new subscribers.

However, the opportunity comes with strings attached. Apple is known for imposing strict guidelines on partners, particularly around data privacy and user experience standards. AI providers looking to integrate with iOS 27 will likely need to meet Apple's privacy requirements, which could include on-device processing mandates, data minimization rules, and transparency obligations.

For app developers, the multi-model approach creates both opportunities and challenges:

• Opportunity: Developers can build apps that leverage whichever AI model the user prefers, reducing dependency on a single provider
• Challenge: Testing and optimizing across multiple AI backends adds complexity
• Opportunity: Users who already subscribe to premium AI services through Apple may be more willing to use AI-powered app features
• Challenge: Developers must handle varying capabilities and response formats across different models
• Opportunity: Apple's platform approach could standardize AI APIs, making integration easier over time

Industry Context: The Rise of AI Aggregation

Apple's move reflects a broader industry trend toward AI aggregation. Companies like Perplexity already allow users to choose between models for search queries. Amazon Bedrock Adds Claude 4 and Llama 4 Models">Amazon Bedrock offers enterprises access to multiple foundation models through a single API. OpenRouter has built a business around routing requests to the optimal model for each task.

What makes Apple's approach unique is scale and integration depth. While existing aggregation platforms serve developers and power users, Apple would bring model choice to mainstream consumers — potentially hundreds of millions of people who have never heard of Anthropic or compared LLM benchmarks.

This democratization of AI model choice could accelerate competition among AI providers. When consumers can easily switch between models, providers must compete on quality, speed, and price rather than relying on distribution advantages. The result could be faster innovation and lower prices across the industry.

Compared to the current landscape where most consumers default to whichever AI comes pre-installed on their device, Apple's approach introduces genuine consumer choice into the AI market for the first time at massive scale.

Looking Ahead: Fall 2025 and Beyond

Apple is expected to unveil the full details of its multi-model AI strategy at WWDC 2025, likely in June. The iOS 27, iPadOS 27, and macOS 27 updates would then ship to consumers in September or October, following Apple's typical release cadence.

Several key questions remain unanswered. Which AI providers beyond Google and Anthropic will be included at launch? Will OpenAI — which already powers some Apple Intelligence features through a ChatGPT integration — expand its role or be displaced? How will Apple handle pricing, with some models offering free tiers and others requiring subscriptions?

The competitive response will also be worth watching. Google may reconsider its exclusive Android integrations if Apple offers Gemini equal footing alongside competitors. Microsoft could push harder to integrate Copilot into hardware partnerships. And AI startups may see Apple's platform as their best path to reaching mainstream consumers.

One thing is clear: Apple's decision to become an AI platform rather than solely an AI provider could reshape how billions of people interact with artificial intelligence. By putting the choice in users' hands, Apple is making a bet that openness — carefully managed openness, in true Apple fashion — will prove more valuable than control in the age of AI.