Apple has banned its employees from using ChatGPT and other AI platforms over fears its own AI plans could be leaked.
According to, ironically, an internal leaked document obtained by the Wall Street Journal, Apple has restricted all use of ChatGPT, Google Bard, and similar large language models (LLMs) for employees whilst it develops similar technology.
Apple employees have also been advised to not use Microsoft’s GitHub Copilot, which developers can use to automate code writing, over fears of leaked confidential data.
This leak comes just as OpenAI released the first official ChatGPT app for the iPhone. The app follows the same model as the web version, with the GPT-3.5 version being free to use. However, those wanting the best version of the technology can upgrade to ChatGPT Plus for $20 a month. The paid-for version allows users to chat with the AI, answer follow-up questions, and even challenge incorrect assumptions.
In a press release to mark the launch of the new iOS app, OpenAI wrote: "The ChatGPT app is free to use and syncs your history across devices. It also integrates Whisper, our open-source speech-recognition system, enabling voice input.
"ChatGPT Plus subscribers get exclusive access to GPT-4’s capabilities, early access to features, and faster response times on iOS.” It also added that the app rollout is in the United States only at present but will expand to additional countries in the coming weeks.
There may be some substance to Apple's concerns. ChatGPT, a Microsoft-supported platform, collects data from users to improve its AI models. But a bug discovered in March allowed users to view the chat history of others, prompting ChatGPT to add the option for users to turn off their chat history and not contribute to AI training.
Obviously, if Apple engineers use ChatGPT, there is no guarantee their codes will not leak or be used by third-party developers.
And it’s not just Apple that has banned its workforce from using AI tech. Companies including JPMorgan Chase, and Verizon have also stopped the use of generative AIs by their employees.
Siri tipped for AI infusion
While Apple seems slow to enter the AI race, a recent discovery by 9to5Mac.com shows it is working on its own AI project. The project nicknamed “Bobcat” has been discovered in the tvOS 16.4 beta and shows Apple testing out new language capabilities for Siri.
The project, which is allegedly being led by John Giannandrea (who Apple hired from Google in 2018) aims to make Siri better understand users and respond more naturally.
So, while current plans don’t seem to include AI technology to rival the likes of ChatGPT, they do want to make Siri more conversational and allow it to perform tasks outside of the narrow conditions that it’s currently been programmed for.
It’s also important to note that under Giannandrea, Apple has acquired a number of artificial intelligence startups. However, it’s unclear when and how Apple plan to make its current AI developments available to the public.
More from Tom's Guide
