Earlier this year, a team of researchers from Wiz Research announced that they had discovered a publicly accessible server belonging to DeepSeek, which contained over a million records. According to them, chat histories, secret keys, and other confidential data were found on the server.
For most of us, the consequences of a data leak depend on what we shared with the neural network. If you discussed something mundane with the chatbot—such as recipes, vacation ideas, or abstract questions—there's not much to worry about. But if personal data, work projects, or confidential information appeared in the chats, the risks become much more serious.
Timofey Voronin, Deputy Director for Technology Transfer at the NTI Competence Center, agrees with him. He believes that even if attackers manage to gain access to the leaked data before it is deleted (according to DeepSeek, the data was deleted within an hour), it will be difficult for them to use it to commit crimes if it only contains culinary recipes and equation solutions.
Pudovkin noted that one of the most serious consequences of a user's data leak could be the use of this personal information to create more sophisticated schemes for deceiving victims using social engineering.
For example, if hobbies, friends' names, or places of work were mentioned in chats, this could become the basis for phishing emails or calls. Information from chats can help attackers create a more accurate "victim profile."
The expert especially highlighted the risks associated with the leakage of confidential information relating to the activities of companies. He emphasized that the loss of work materials, passwords, or financial data could have serious consequences.
The most serious consequences of such leaks can be for those who used DeepSeek not for personal, but for work purposes. It is not surprising that many companies now directly prohibit the use of artificial intelligence to solve official tasks. For example, for writing code.
DeepSeek is a neural network created by the Chinese company of the same name. It was presented to the public in January of this year.
Earlier, Kaspersky Lab discovered a new virus that penetrates smartphones through fake applications in the App Store and Google Play. This Trojan steals data from photos and can be found in 20 different programs: from messengers to AI assistants and crypto exchanges.
Read more on the topic:
The cost of vulnerabilities: the income of "white hackers" increased significantly in 2024 in Russia
Hackers are hacking smart devices in homes and forcing them to "work" for them
Your voice is familiar to us: telephone scammers will begin to be identified using biometric data
Mass iPhone hacking: scammers have come up with a new type of fraud