A chatbot with elements of artificial intelligence ChatGPT has leaked corporate data of employees of the South Korean company Samsung. These release reports Economist with reference to sources.
It was recorded that the employees of the South Korean IT giant started using the chatbot in their work about three weeks ago, but they acted carelessly, for example, they talked about developments in the field of processors, new gadgets and their functionality long ago. official announcement.
“Enterprise experts abused ChatGPT because Samsung’s confidential data was freely accessible,” said a source inside the company.
In one case, an engineer entered the source code for chip-making equipment into an AI chatbot thread, according to The Economist. The second known leak occurred when another person shared secret source code with ChatGPT to facilitate verification.
The third leak occurred when a Samsung employee tried to create a text version of the meeting minutes. In this case, the information hidden to people outside the company has become a part of the knowledge base of artificial intelligence.
According to the publication, Samsung has taken measures to prevent the recurrence of leaks. Employees were asked not to share any confidential information and data of the South Korean seller with the chatbot.
Former Wikipedia founder Jimmy Wales said about plans to pull ChatGPT from OpenAI to write articles in an online encyclopedia in the near future.