AI-powered telemarketing robots widely used for unsolicited calls in China
AI-powered telemarketing robots in China are making hundreds of unsolicited calls daily using spoofed caller IDs, targeting consumers indiscriminately without prior consent. Newer large language model systems can generate personalized scripts and deceive users about whether they are speaking with a human, exacerbating harassment concerns as costs have dropped enough for widespread deployment.
What happened
AI-powered telemarketing robots are making 800 to 1500 outbound calls per day in China, deployed through technology companies that charge by the minute (0.12 to 0.16 yuan per minute). A new generation of large language model-based systems can automatically generate conversation scripts without pre-written text and claim to provide personalized communication through semantic understanding and emotion recognition.
The AI robots use number spoofing techniques to display local area codes and appear as regular mobile or landline numbers, making it difficult for consumers to identify the calls as automated. The industry has capabilities to remove spam call labels from numbers at a frequency of up to three times per day.
Reported impact
Consumers report a significant increase in unwanted marketing calls covering real estate, education, finance and insurance sectors during the summer months. Shanghai Consumer Council research found that AI calls often employ indiscriminate 'carpet bombing' dialing, contacting people who have never registered for related services by automatically generating phone numbers without regard to time of day or target selection.
Index analysis
Beijing University of Posts and Telecommunications professor Cao Jianqiu noted that new AI telemarketing systems have strong identity concealment, can mimic human characteristics in calls and sales pitches, and can understand and answer questions as flexibly as humans, making it difficult for consumers to identify whether they are speaking with a real person or a machine. The dramatic cost reduction of AI calling (from previous availability only to high-profit industries to now affordable for all sectors) combined with technological advances in deception capability have increased the volume and sophistication of unsolicited calls.
According to China's Consumer Protection Law Implementation Regulations Article 24, operators must not send commercial information or make commercial calls to consumers without their explicit prior consent. Legal experts identify the boundary between compliant marketing and harassing calls as whether the user has clearly consented in advance, and recommend that AI calls should be limited to scenarios such as after-sales notifications, performance reminders, and customer follow-ups where authorization has been obtained, with strict restrictions on cold commercial solicitation.
Governance controls
Some e-commerce platforms have begun restricting searches for keywords such as 'AI robot' and 'intelligent outbound calling,' with results redirected to error pages or replaced with policy and public welfare information. Consumers have two formal feedback channels: reports to the 12321 Network Misconduct and Spam Reporting Center, and free enrollment in 'harassment call rejection' protection services offered by the three telecommunications operators using technical interception.
Legal experts recommend that future regulation should strengthen call line registration and phone number authentication to combat indiscriminate 'carpet bombing,' clarify responsibility chains among technology service providers, deploying companies, and telecommunications operators, narrow application scenarios for AI outbound calling, and mandate built-in technical compliance such as mandatory do-not-call blacklists, call frequency limits, time-of-day controls, and clear identification of AI-generated calls.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.