Apple reportedly testing its own generative AI chatbot, but we might never see it

With the rise of OpenAI’s ChatGPT and Google Bard, Apple is also working on its own “Apple GPT” AI chatbot, according to Bloomberg‘s Mark Gurman. The publication claims Apple doesn’t have a plan to release the technology to consumers yet. However, it’s already using this AI chatbot internally to develop prototypes and improve existing products.

According to people with knowledge of the efforts, several teams are collaborating on the large language model framework, known as Ajax, which powers the chatbot. Currently, Apple’s best effort regarding AI is Siri, but previous reports have revealed how difficult it is to make improvements to the personal assistant due to the way it was built.

With the upcoming iOS 17, Apple has announced AI-first initiatives that it calls machine learning improvements, such as a better language model for autocorrection, removing “Hey” from “Hey Siri,” revamps to spotlights, making it easier to find photos, and more.

Gurman says that although Apple is trying to publicly ignore reports about it losing the AI battle, behind the scenes, the company has “grown concerned about missing a potentially paramount shift in how devices operate” as “generative AI promises to transform how people interact with phones, computers, and other technologies.”

The journalist reports that Search, Siri, and Maps received deployed AI-related improvements with Apple GPT, and this feature is being used to create large language models and serve as the foundation for an internal ChatGPT-style tool.

The tool reportedly replicates Bard, ChatGPT, and Bing Chat, but offers no breakthrough features. While Apple still tries to decide how it’s going to tackle generative AI, the company has been hiring engineers in that field and using this tool internally to develop improved LLM technology.

BGR will keep reporting on this matter as we learn more about Apple’s plans.