Samsung employees accidentally leaked company information on ChatGPT

(Representational Image)

Samsung has reportedly logged three separate instances of employees leaking data, in the twenty days that ChatGPT has been authorised for use at their semiconductor facilities.

As spotted by Mashable, and reported on by The Economist Korea, in one instance, a Samsung employee pasted confidential company source code into the chatbot to check for errors.

Also Read | Samsung to cut chip production after 14-year low profit

Another employee also shared code with the AI chatbot, requesting for "code optimization". In the third instance, an employee shared a recording of a confidential company meeting to convert into notes.

Since there are no secrets online, that information is now part of ChatGPT forever, and this has no doubt annoyed Samsung.

Interestingly, there is a lot of increasing scrutiny over OpenAI and ChatGPT's data collection policies in Europe. Italy has outright banned the AI chatbot over privacy concerns.

As reported, Samsung has now limited ChatGPT's uploads to only 1024 bytes per person and is also investigating the employees that were part of the leak.

Also Read | Samsung Galaxy S23 Review: Plenty of smartphone in a pocket-sized package

Samsung is also reportedly considering building its own version of the AI chatbot, to prevent mishaps like these in the future.

OpenAI has already stated that it uses the data collected by the chatbot to train its AI models. It also warns people not to share any sensitive data with the chatbot.