Apple is set to roll out iOS 27 later this year, and the forthcoming operating system will introduce a capability that could redefine how consumers interact with artificial intelligence on personal devices. According to a Bloomberg report obtained by TechCrunch, the new software will embed a feature internally known as “Extensions,” which will allow users to select from a menu of external large‑language models (LLMs) to power functions such as voice assistance, text generation, and image manipulation. The feature will be mirrored in iPadOS 27 and macOS 27, extending the model‑selection framework across Apple’s entire consumer‑device portfolio.
The proposed architecture positions Apple’s own suite of AI tools—collectively branded as Apple Intelligence, which includes Siri, Writing Tools, and Image Playground—as a thin integration layer that can invoke any compatible model installed on the device. In practice, a user could, for example, ask Siri to draft an email using an Anthropic Claude model, while a separate app might call a Google Gemini model for image captioning, all without leaving the native interface. The report notes that Apple is already testing models from Google and Anthropic, while the status of OpenAI’s ChatGPT remains uncertain. Because ChatGPT currently powers Apple’s on‑device AI features, its continued inclusion would likely be a default option for many users.
The timing of the announcement coincides with a leadership transition at Apple’s helm. Tim Cook, who has guided the company for more than a decade, is slated to step down in the coming months, and senior vice president John Ternus—who has overseen hardware engineering and product design—will assume the chief executive role. Ternus inherits a strategic crossroads: while Apple’s rivals have been launching stand‑alone AI services and cloud‑based platforms, the Cupertino giant has largely kept its AI ambitions under wraps, focusing instead on embedding intelligence directly into its devices.
Industry analysts view the move as an attempt to close a perceived gap with competitors such as Microsoft, which integrates OpenAI’s models across its Windows operating system and Office suite, and Google, which embeds Gemini into Android and its own suite of productivity tools. By allowing third‑party models to run on‑device, Apple sidesteps the need for massive data‑center investments and aligns with its long‑standing privacy‑first narrative. The approach also leverages Apple’s premium hardware, which already includes the Neural Engine—a custom silicon component designed for machine‑learning workloads. Turning the iPhone, iPad and Mac into flexible AI endpoints could unlock new revenue streams from developers eager to monetize model‑specific extensions, while preserving the company’s control over the user experience.
From a market perspective, the shift could have several implications. First, it may broaden the addressable market for AI model providers. Companies such as Anthropic, which raised $4 billion in 2024 and is valued at roughly $30 billion, could see a surge in demand for model licenses tailored to Apple’s ecosystem. Google’s Gemini, already a cornerstone of its own cloud AI offerings, would gain a foothold on a platform that commands roughly 24 percent of global smartphone shipments, according to IDC data for Q4 2025. If Apple makes model selection a seamless part of everyday workflows, the resulting data flows could inform future pricing models and partnership structures.
Second, the move reinforces Apple’s hardware‑centric revenue model at a time when the broader technology sector is grappling with slowing device sales. In its fiscal 2025 results, Apple reported $115 billion in services revenue, a 12 percent year‑over‑year increase, but hardware still accounted for nearly 80 percent of total earnings. By embedding AI capabilities directly into devices, Apple can differentiate its premium hardware without relying on external subscription services, a strategy that may appeal to markets where data‑privacy regulations are tightening.
Geopolitically, the development arrives amid intensifying US‑China competition over AI leadership. The United States has recently enacted the AI Innovation and Competition Act, which allocates $25 billion for domestic AI research and incentivizes private‑sector collaboration. At the same time, China’s New Generation AI Development Plan targets a 70 percent global share of AI patents by 2030. Apple’s decision to support models from both American firms (OpenAI, Anthropic) and a multinational like Google—while keeping the execution on US‑designed silicon—underscores a strategic alignment with Washington’s push for secure, on‑device AI that limits dependence on foreign cloud infrastructure. The approach also mitigates exposure to export‑control restrictions that could affect cloud‑based AI services, a concern that has surfaced in recent trade talks between the two powers.
The broader AI market is projected by IDC to exceed $1.5 trillion in annual spend by 2027, with generative AI accounting for a growing slice of enterprise and consumer budgets. Apple’s move to democratize model access on its devices could accelerate consumer adoption, especially in regions where mobile devices remain the primary computing platform. For developers, the “Extensions” framework may resemble a new app store tier, where model providers and software creators bundle specialized AI capabilities for niche use cases—ranging from multilingual translation to real‑time video editing.
While Apple has not disclosed specific timelines beyond the “later this year” rollout window, the company’s internal testing suggests a relatively mature implementation. The firm’s emphasis on privacy, combined with its control over the hardware stack, positions it uniquely to offer a differentiated AI experience that does not rely on continuous cloud connectivity. As the industry watches the transition from Tim Cook to John Ternus, the success of iOS 27’s multi‑model approach could serve as a litmus test for Apple’s ability to pivot from a hardware‑only narrative to a hybrid model that blends device excellence with flexible, third‑party AI services.
In sum, the forthcoming iOS 27 update signals a strategic recalibration for Apple, one that could reshape competitive dynamics in the generative AI arena, reinforce the company’s hardware‑first philosophy, and align its roadmap with broader geopolitical currents favoring secure, on‑device intelligence. The rollout will be closely monitored by investors, policymakers, and rival tech firms alike, as it may set a precedent for how consumer electronics manufacturers integrate external AI models without ceding control of the user experience.