Samsung Electronics temporarily limits in-office use of generative AI

Samsung Electronics Co. headquarters in Seoul [Photo by Yonhap]

Samsung Electronics Co. has decided to restrict the use of generative artificial intelligence tools, including the recently popular ChatGPT, at some parts of its offices on a temporary basis on concerns of potential leakage of confidential information.

The restriction has been applied in the company’s Device Experience (DX) division, which covers the mobile phone and home appliance businesses, since the company announcement at the end of April, according to the company on Tuesday.

Following the restriction, no staff members or executives in the DX division are allowed to use generative AI tools within the intranet network, nor on any other computers owned by the company.

Samsung Electronics left some room, however, to roll back the restriction in the future on condition that certain security measures are in place, considering that these tools might be beneficial to work in terms of productivity or efficiency.

Earlier, the company’s Device Solution (DS) division, which oversees the chip business, also announced a warning about the unlimited use of generative AI tools, as it faced leaks of certain source codes, triggering an internal debate concerning security and the use of generative AI. To this end, the company conducted an in-house survey to which 65 percent of all respondents, including staff members and executives, answered that those AI tools pose a security risk.

The DS division has not blocked the use of generative AI completely and mostly called for attention to be paid to any misuse or abuse of the new technology. The division also conducted a training session on the scope of the use of generative AI.

“The limitations are being imposed temporarily until we can come up with countermeasures,” said one Samsung Electronics official. “Eventually, we aim to allow the use of generative AI in a secure environment to improve work efficiency.”

The electronics giant also reportedly urged its employees to be cautious when using generative AI not only on office computers but also outside the office, making sure not to enter any company-related information or personal information into the chatbot.

Amid growing interest in generative AI around the world, some companies have completely banned the use of generative AI. In February, some large U.S. financial institutions, such as JPMorgan Chase & Co., Bank of America Corp. and Citigroup Inc. restricted the use of generative AI, including ChatGPT by Open AI, Bard by Google LLC and Bing by Microsoft Corp. Those companies are understood to have put in place strong measures as they believe data sent to those AI platforms can be stored on external servers, making it difficult to manage data and raising the risk of leaks.