Apple Explores Third-Party AI Models For Siri in Bid to Catch up in AI Race

Tech giant Apple is considering integrating Anthropic’s Claude or OpenAI’s ChatGPT to enhance its Siri assistant, according to Bloomberg reports.

While Apple has traditionally relied on its own AI technology, it has reportedly engaged both AI companies to train large language models for testing on its cloud infrastructure. This move comes as Apple’s AI-upgraded Siri, initially planned for 2025, has been delayed to 2026 due to technical setbacks, highlighting the company’s struggle to keep pace in the competitive AI landscape.

Apple is perceived to be lagging in the AI race. Its AI efforts, including Siri, have faced challenges, with critics noting that Siri has not kept pace with competitors like Google Assistant, Amazon’s Alexa, or advanced models like ChatGPT. Recent reports highlight internal struggles, such as slower progress in developing proprietary large language models and a management shakeup, including the reassignment of AI chief John Giannandrea.

Apple’s exploration of third-party AI models from OpenAI or Anthropic for Siri suggests an acknowledgment of these gaps, as it seeks to bolster its AI capabilities to remain competitive. However, Apple’s strong ecosystem and integration potential could still position it to catch up if it leverages external partnerships effectively.

The tech giant continues to develop an internal project called “LLM Siri,” which aims to power a revamped Siri using in-house models. However, the company is now exploring deeper integration with third-party AI technologies, building on Siri’s existing ability to leverage ChatGPT for complex queries.

This potential pivot to external models signals Apple’s acknowledgment of its challenges in generative AI, a critical technology where it trails competitors like Google, OpenAI, and Anthropic.

Suff Syed, a Product & Design leader at Microsoft Research, commented on the development, stating,

“Apple is behind and the market knows it. Apple Intelligence and Siri are a punch line. But the bigger issue is Apple renting intelligence from the very companies out-innovating it. If it doesn’t do anything radical, it’s going to end up like IBM/HP. Rich, but irrelevant.”

Technology journalist for Bloomberg Mark Gurman in a post on X, disclosed that Apple’s effort to explore outside large language models instead of its foundation models has taken a toll on its AI team. Tom Gunter, one of its top engineers, left last week. And the team behind MLX, Apple’s open-source AI framework has threatened to quit.

Despite these concerns, Apple has not fully abandoned its in-house efforts, and it is concurrently running a project called LLM Siri that would power next year’s Siri overhaul with in-house AI. Top executives reportedly see third-party models as a potential way to close the gap with rivals.

Apple’s shares climbed 2% on the news that the tech giant is considering integrating third-party AI models for Siri. The surge in shares signals investor optimism about the potential integration of third-party AI models from OpenAI or Anthropic into Siri, which could enhance its capabilities and competitiveness. It further reflects market confidence in Apple’s strategy to modernize Siri and strengthen its position in the AI race, despite challenges in its in-house AI development.

The revamped Siri, expected with iOS 26.4 in 2026, aims to deliver advanced capabilities, including improved context-awareness and the ability to handle complex, multi-step commands by leveraging on-screen content and user data. While Apple weighs its options, no final decision has been made on abandoning its proprietary AI models.