- Apple is working on its own AI large language model, the WSJ reported.
- Its efforts are being led by Google's former AI chief and Tim Cook says it plans to "weave" AI into products.
Apple is working on its own AI and restricting its staff's ChatGPT use over privacy concerns, The Wall Street Journal reported.
The tech giant is building a large language model – an AI system similar to the likes of ChatGPT and Google's Bard – people familiar with the matter told the Journal.
Its efforts are being led by John Giannandrea, Google's former AI chief who joined Apple in 2018.
In 2020, Apple then purchased two AI startups for $200 million and $50 million. Its voice assistant Siri also uses machine learning to interpret what users are saying.
In the company's quarterly earnings call earlier this month, CEO Tim Cook said Apple is planning to "weave" AI into its products, per The Independent. But he also cautioned about the future of the technology.
"I do think it's very important to be deliberate and thoughtful in how you approach these things," he said, per Inc. "And there's a number of issues that need to be sorted as is being talked about in a number of different places, but the potential is certainly very interesting."
Apple is also telling some employees to limit their use of ChatGPT and other external AI tools, according to an internal document seen by the Journal. That includes the automated coding tool Copilot, from the Microsoft-owned GitHub.
The Journal's document reportedly says Apple is concerned that staff could release confidential data. That's because users' inputs are sent back to the developers to help train the AI models.
Last month, ChatGPT introduced the ability to turn off chat history to give users more privacy controls. But Amazon and Samsung have also restricted their staff's use of the chat bot.
On Thursday, OpenAI announced its official ChatGPT app for iPhone users in the US.
Apple did not immediately respond to Insider's request for comment.