Apple reportedly using Gemini to test ChatGPT-style answers

268

Synopsis:

Apple is quietly testing Google’s Gemini AI model to power enhanced chatbot-style answers in iOS, signaling deeper interest in LLM integration.

Summary:

Apple is reportedly experimenting with Google’s Gemini AI model to test ChatGPT-style responses within iOS, according to a report by The Information. The move suggests Apple may be open to working with third-party large language models (LLMs) as it explores more advanced AI capabilities across its ecosystem.

Sources familiar with the matter said Apple engineers have been evaluating Gemini’s performance, particularly in delivering conversational answers to user queries. These internal tests do not currently reflect a commercial deal, but the evaluation hints at Apple’s increasing interest in expanding its generative AI footprint.

This development follows growing speculation that Apple may partner with leading AI companies to deliver generative experiences inside iPhone and MacOS interfaces. Apple’s current on-device AI capabilities — such as Siri and text suggestions — have been seen as lagging behind more advanced LLMs like OpenAI’s GPT-4 or Google’s Gemini.

While Apple is developing its own LLMs under a project internally called Ajax, the company is reportedly weighing multiple integration paths, including potential collaborations with other tech giants. That includes Google, OpenAI, and potentially even Anthropic — all of which are already in talks with various device manufacturers and platforms.

If formalized, a Gemini-Apple partnership could mark a shift in Apple’s historically closed AI development approach, similar to how it incorporated Google Search as the default engine in Safari. It could also give Google’s Gemini a crucial edge in consumer adoption, particularly if embedded within one of the world’s most widely used operating systems.

Neither Apple nor Google has commented officially on the development.