Deepfake technology, which uses AI's deep learning algorithms to simulate and forge images, audio and video, can be misused or used maliciously to endanger individuals' legal rights, social stability and national security, the Ministry of State Security (MSS) said in an article on Friday.
The newspaper said some “hostile foreign anti-China forces” had used the technology to generate fake videos and tried to spread them within China to “mislead public opinion and cause panic.”
The ministry added that the use of AI could lead to data breaches and security risks. The report cited instances where staff used open source AI tools that allowed users to view files without passwords, allowing anyone with an external IP address to access and download sensitive data.
The department also warned about algorithmic bias, which means large language models can amplify discrimination if training data reflects social bias or underestimates certain problems.
11:56
From India to China, how deepfakes are reshaping politics in Asia
From India to China, how deepfakes are reshaping politics in Asia
To combat these risks, the ministry said users should be careful when granting permissions, avoid using web-based AI to process sensitive data, avoid storing payment passwords, and disable data sharing or cloud storage. They should also monitor their “digital footprint,” regularly clear their AI chat history, update their antivirus software, and be wary when asked to provide personally identifying information or bank account details.
