A month after internal, sensitive data belonged to Samsung accidentally leaked to ChatGPT, Samsung is cracking down on the use of the generative AI service. The electronics giant is planning a temporary ban on the use of generative AI tools on company-owned devices, including computers, tablets, phones and non-company-owned devices that run on internal networks. The ban would not only cover ChatGPT, but also services that use the technology, such as Microsoft’s Bing, and rival generative AI services such as Google’s Bard.
The impending ban was initially reported by Bloomberg. The rule would only apply to devices Samsung issues to its employees, meaning consumers and others who own Samsung phones, laptops and other connected devices would not be affected. It is not clear if it is already in effect and if not when it will come into effect.
Samsung did not immediately respond to a request for comment.
According to a memo seen Monday by Bloomberg, the restriction would be temporary and permanent until “security measures are put in place to create a safe environment for using generative AI safely to improve employee productivity and efficiency.” The South Korea-based tech company is said to be developing its own in-house AI tools for “software development and translation,” according to the report.
Open AI’s generative AI chatbot ChatGPT has exploded in popularity worldwide since its launch last November, with people leaning on it to give them text-based answers to everything from basic research to business tasks and more. But some of that AI boom is running into significant roadblocks. In addition to leaking proprietary data to the service, like last month with Samsung, others have flagged potential data privacy violations, copyright violations, and inaccuracies in ChatGPT’s responses.
The tech giant initially allowed employees of its Device Solutions (DS) division, which manages its semiconductor and display business, to use generative AI starting March 11. In the wake of the data breach, Samsung also asked staff using generative AI tools elsewhere to submit company-related information or personal data “that could reveal its intellectual property,” according to the memo reviewed by Bloomberg.
One of the issues Samsung noted is that it is difficult to “get and delete” the data on remote servers, and the data sent to such AI tools could be disclosed to other users. Based on Samsung’s internal survey in April, about 65% of participants said using generative AI tools poses a security risk.
OpenAI has worked to fix some of the more controversial issues to lift some of the more high-profile bans. Recently, ChatGPT services resumed in Italy after OpenAI unveiled a plan to introduce new privacy controls. Big banksincluding Bank of America, Citi, Deutsche Bank, Goldman Sachs, Wells Fargo and J.P. Morganare among the other companies that have recently restricted employees’ use of ChatGPT.
In South Korea, this includes other major tech companies, including LG and memory chip maker SK Hynix struggle to make their own guidelines for using generative AI tools.
