Login to Continue Learning
The next-generation of Siri, set to arrive in 2026, is expected to feature new capabilities that will put Apple back on the AI map. However, Apple faces challenges deciding whether to develop its own Large Language Model (LLM) or rely on a partner. After discussions with OpenAI and Anthropic, reports indicate Apple is now in talks with Google about using its Gemini AI model for the updated version of Siri.
Apple is currently testing multiple LLMs to determine which option will deliver the best results to customers. Bloomberg reported that Apple approached Google to develop a custom AI model for the new version of Siri. In response, MacRumors noted that Google has begun training a model intended to run on Apple’s Private Cloud Compute servers.
So far, in the U.S., Apple has discussed partnerships with three entities regarding using third-party LLMs for the revamped version of Siri. Deciding which option to pursue is complex; Apple must choose between developing its own LLM, which would be costly and uncertain, or relying on companies like Google. Historically, Apple has been late to adopt new technological standards, but in this case, it faces significant criticism for its AI failures, including a lawsuit over false advertising of Siri’s capabilities.
Apple aims to excel in this area and is testing several LLMs. The process will likely take some time, but updates on the company’s progress will be provided. For now, Apple is exploring options with multiple partners before making a final decision.
News Source: Bloomberg
