AI Mobile Assistant Safety Concerns Spark Discussion at Two Sessions; Douyin VP Li Liang Extends Invitation to Committee Members for Product Testing to Address Doubts
2 day ago / Read about 0 minute
Author:小编   

At the 2026 National Two Sessions, the safety and governance of AI technology emerged as a focal point of debate. Li Mengjiao, a member of the National Committee of the Chinese People's Political Consultative Conference (CPPCC), highlighted the potential safety hazards associated with current mobile AI assistants, including the excessive gathering of personal information and ambiguous permission boundaries. She cited instances where certain AI assistants, capable of "taking over the phone," exploit "visual screen reading" and "simulated clicking" technologies to access low-level system permissions. This enables them to carry out multiple operations without the user's knowledge, potentially exposing sensitive information such as bank verification codes. Such capabilities have given rise to the "digital god" phenomenon, where AI assistants possess a more comprehensive understanding of users' digital footprints than the users themselves, posing a threat to personal privacy and financial security. Li Mengjiao urged for expedited enhancements to the relevant safety management frameworks.

In response, Li Liang, Vice President of Douyin Group, suggested that there might be misconceptions at play. He clarified that all operations of legitimate mobile AI assistants necessitate active initiation by the user, with sensitive processes being relinquished to the user for handling. These assistants do not autonomously execute high-risk instructions without the user's awareness. Earlier, Doubao Mobile Assistant had also issued a statement addressing rumors of "security vulnerabilities," asserting that the authors had maliciously disseminated and exaggerated the risks associated with these vulnerabilities. Furthermore, the unauthorized disclosure of such vulnerabilities could potentially constitute a violation of the law. Li Liang underscored the importance of differentiating between legitimate technical operations and malicious attack behaviors, emphasizing that irresponsible disclosure methods of this nature may breach relevant national regulations.