Efforts ramped up against AI-enabled crimes

Chinese police have intensified measures to combat crimes involving artificial intelligence, including fraud, identity forgery, data theft and the creation of online rumors, a senior Ministry of Public Security official said on Friday.

Individuals are increasingly using AI and other emerging technologies to commit telecom and online fraud, violate personal privacy, undermine public trust and, in some cases, threaten national security, Jiang Guoli, director of the ministry's Criminal Investigation Bureau, told a State Council Information Office news conference.

He said AI-enabled crimes mainly involve using the technology to forge identities, break into government or corporate information systems, and steal personal information or property.

Some criminals use AI-generated images, voices or videos to impersonate acquaintances or authoritative figures and gain victims' trust before committing fraud. Others use AI tools to fabricate and spread online rumors, disrupting public order, according to him.

Jiang said police are targeting both those who commit such offenses and criminal groups that provide technical support, revealing that in the first half of this year, police investigated more than 170 cases involving online rumors generated with AI tools and dealt with over 190 people.

Authorities have also developed risk assessment models using big data to identify suspicious activity, including frequent video calls from unknown accounts, calls involving silent audio recording, repeated attempts to obtain facial images, and video requests initiated from overseas, he said.

While warnings and case leads generated by the models are sent to local police to help prevent potential victims from transferring money or disclosing personal information, the National Anti-Fraud Center app has also introduced an AI-content identification function, where users can upload suspicious videos, audio recordings or images to check whether they contain signs of AI generation or manipulation, he added.

The ministry has worked with other authorities to issue rules requiring labels for AI-generated and synthetic content, he said. Regulators have also strengthened inspections of AI services and monitored potential security risks associated with large language models.

From January to June, authorities conducted more than 600 inspections of AI application services and took over 30 administrative enforcement actions, including issuing warnings, ordering corrections and shutting down services.

Jiang said authorities are also advancing legislation on cybercrime prevention and developing security review mechanisms and technical standards for AI-generated code, deepfake images and other applications.