Samsung bans staff use of ChatGPT after confidential code leaks
The ban followed engineers pasting proprietary semiconductor source code and a confidential meeting transcript into the chatbot within weeks of Samsung lifting an earlier internal restriction.
- Security & misuse
- Minor
Samsung Electronics banned employee use of ChatGPT and other generative AI chatbots after engineers in its semiconductor division uploaded sensitive internal material to the service. Reporting at the time described several incidents, including source code for chip-equipment diagnostics and a transcript of an internal meeting being pasted into ChatGPT prompts, in the weeks after Samsung had lifted an earlier internal restriction and allowed engineers to use the tool.
Samsung’s stated concern was that data submitted to ChatGPT is stored on servers run by OpenAI, Microsoft and other cloud providers, with no straightforward way for the company to retrieve or delete it, and that such data could in principle surface in outputs served to other users. The company issued a memo prohibiting staff use of generative AI tools generally, rather than singling out ChatGPT alone.
The episode became one of the most widely cited early examples of the data-leakage risk consumer AI chatbots posed to corporate users, at a moment when many large companies were still deciding whether to permit employee use of such tools at all. It contributed to a wave of enterprise bans and policies restricting staff use of public AI chatbots through 2023, and to renewed interest in walled-off, enterprise-grade alternatives that kept prompt data off third-party servers.