While we talk about the many benefits of AI technology, we should not ignore the data protection and privacy concerns raised by services that use this technology. After all, artificial intelligence is a technology that feeds on the data provided by users and constantly trains itself. Chatbots, which are very popular nowadays, use the content of each conversation with users for their own training.
It would not be wrong to say that the first thing that comes to mind when it comes to data protection is the protection of personal data, as the laws on personal data protection are strictly used all over the world and individuals are becoming more and more aware of this issue. However, artificial intelligence technology poses at least as much of a threat to companies’ confidential data.
One of the best examples of this is that in April this year, Samsung employees leaked some of the company’s confidential and sensitive information by copying it to the ChatGPT chatbot on three different occasions. In one of these incidents, a Samsung employee copied the confidential source code of a company software program to the ChatGPT chatbot and asked it to find a bug in the source code. When the company management realized these data leaks, they banned Samsung employees from using ChatGPT.
Copying the company’s confidential information in this way means that this information is sent to servers controlled by third parties and out of the control of the data owner. The fact that artificial intelligence, which trains itself with the data it processes, can use this information in its interactions with other users increases the severity of the situation.
In the face of this risk of leakage, companies that want to ensure the security of company data are considering various methods such as banning the use of artificial intelligence-based services and limiting access to these services to certain levels of employees.
In the light of all these developments, Open AI, which developed ChatGPT, has taken a step towards increasing the security of company data. Open AI made an announcement last Monday (August 28, 2008) regarding its new service ChatGPT Enterprise and underlined that one of the most important features of ChatGPT Enterprise is that it will provide security and privacy at the corporate level. According to the privacy policy of Chat GPT Enterprise, companies will have full control over their own data and this data will not be included in the learning cycle of artificial intelligence.
A complete ban on AI-based services does not seem realistic as a long-term solution. However, ensuring the security of companies’ data, which is perhaps their most valuable asset today, is a necessity that cannot be underestimated. There is no doubt that developments in artificial intelligence and data security will continue at a rapid pace.