Apple eyes major Siri upgrade as talks with OpenAI, Anthropic reach advanced stage.
Shift toward external AI models signals new era in Apple’s voice assistant strategy.
Apple is reportedly in advanced discussions with leading artificial intelligence companies Anthropic PBC and OpenAI to integrate their large language models (LLMs) into future versions of Siri. This move signals a strategic shift from the company’s long-standing dependence on internally developed AI systems.
Industry insiders suggest that Apple is exploring the deployment of these third-party AI technologies such as Anthropic’s Claude and OpenAI ChatGPT, on its Private Cloud Compute servers, which are powered by Apple’s custom Mac chips. This approach aims to safeguard user privacy and data security, aligning with Apple’s privacy-first philosophy while potentially enhancing Siri’s long-awaited development.
Since its launch in 2011, Siri was one of the first mainstream virtual assistants but has since lagged behind more sophisticated generative AI solutions. Previously, Apple planned to upgrade Siri using its proprietary Apple Foundation Models in a major update scheduled for 2026. However, persistent engineering delays and mixed performance reviews prompted a reevaluation of this strategy. With Mike Rockwell now overseeing Siri’s engineering and Craig Federighi, Apple’s head of software, directly involved in AI decision-making, internal assessments reportedly indicated that Anthropic LLMs outperformed Apple’s own models during early tests, leading to serious interest in forming an external partnership.
This potential shift has created unease within Apple’s AI divisions. Some engineers have voiced concerns that relying on external technology may be seen as a failure of internal innovation efforts, triggering fears of a talent exodus, especially as competitors like Meta and Google offer significantly higher compensation packages for AI professionals.
This situation may also further marginalize Apple AI chief, John Giannandrea, who strongly advocates for internal AI development. Recently, many AI-related teams have been reassigned to report directly to Federighi, and some in-house projects, such as the AI coding assistant Swift Assist, have already been shelved in favor of integrating third-party models.
A key obstacle to any external partnership is cost. Reports suggest that Anthropic is seeking a multi-billion dollar annual fee, which could increase over time. This pricing may prompt Apple to consider OpenAI as a potentially more economical alternative, although negotiations with both companies are still ongoing. If Apple formalizes a deal, it would represent one of its most significant AI decisions in years, potentially reshaping how Siri competes with voice assistants from Google and Amazon while repositioning Apple in the rapidly growing generative AI market.
Despite exploring third-party LLMs, Apple continues to invest in its own AI models for on-device tasks, developer tools, and services that require deeper hardware-software integration. Sources indicate that the company is contemplating a hybrid AI strategy, combining proprietary solutions with selected external technologies to meet immediate performance goals without compromising long-term control.
Apple’s willingness to enhance Siri with external AI tools marks a significant shift in its AI strategy. As user expectations evolve rapidly and competitors push the boundaries of generative technology, Apple appears increasingly committed to closing the performance gap regardless of the source of innovation.
