Number De-tagging + Bulk AI Outbound SaaS Harassment & Fraud: A Gray Industry Chain Masked as Telemarketing Tools for Fraudulent Lead Generation

Who Gets Targeted

The victim base is extremely broad, ranging from ordinary consumers to small and micro-business owners. Ordinary users, fatigued by frequent AI harassment calls, often lower their guard and are deceived by AI scripts impersonating e-commerce, finance, education, or real estate customer service, leading them to fake support agents or phishing refund links. Small and micro-business owners are often lured by telemarketing agents promising thousands of calls per day, high ROI, and precise number pools. However, these pools consist of illegally obtained personal contact information, and the 'de-tagging' services are operated through gray market channels. Ultimately, business owners not only lose their investment but also face legal penalties for involvement in illegal activities. Their psychological vulnerabilities lie in the blind belief in efficiency tools, numbness toward harassment calls, and a sense of desperation for customer acquisition.

骗局怎么运作

• Step 1: Customer acquisition via tool promotion. Fraud rings register shell companies and advertise AI smart telemarketing SaaS on platforms like Baidu, Douyin, and Qichacha. They claim to offer thousands of calls per day, reduce acquisition costs to 3 yuan, and bypass carrier blocking, attracting business owners and telemarketing teams with tiered pricing from monthly to quarterly packages.
• Step 2: Illegal supply of number pools. Operators acquire leaked personal phone numbers through black market channels or use web crawlers to scrape contact information from public B2B platforms. They build databases of millions of numbers categorized by industry. Once a client pays, they can select target demographics in the backend and launch AI outbound tasks with one click, using sources that are non-compliant and unauthorized.
• Step 3: Standardized AI scripts. The platform provides pre-set industry scripts generated by large models to create human-like dialogues. These can adjust branches in real-time based on the recipient's responses, simulating human tone and pauses. Some versions even switch regional dialects, making it difficult for recipients to identify the call as machine-generated, thereby increasing call duration and trust to pave the way for human hand-offs or WeChat additions.
• Step 4: Number de-tagging and spoofing. To address the issue of low connection rates due to harassment tags, the platform provides de-tagging services by submitting bulk false-positive appeals to providers like Tencent Number Assistant or TouchPal. Simultaneously, they use spoofing software to constantly rotate outbound caller IDs, making it difficult for carriers and anti-fraud systems to identify and block them, creating a technical closed loop that bypasses regulation.
• Step 5: Monetization through lead generation. For consumers, the outbound call serves as an entry point to impersonate customer service regarding refunds, prizes, or expired memberships, inducing them to add WeChat or click links for fraud. For business clients, the platform maintains subscriptions by forging call data, inflating connection rates, and faking call logs. Once a client complains or requests a refund, the platform blocks them and disappears, creating a dual-harvesting model.

红旗信号（看到这些快跑）

• 🚩 Telemarketing SaaS ads claiming over 1,000 calls per day with promised ROI or acquisition costs below 5 yuan, which often rely on illegal number pools and AI human-like scripts.
• 🚩 Outbound platforms that offer 'number de-tagging' services or claim to bypass carrier blocking mechanisms are clear indicators of gray market operations.
• 🚩 Service contracts that require business clients to sign liability waivers or hidden clauses that shift the responsibility for the source of the numbers to the buyer.
• 🚩 Platform demo scripts that include typical fraudulent lead generation tactics, such as impersonating customer service, creating a sense of urgency, or guiding users to add WeChat.
• 🚩 Payment models that require large upfront quarterly or annual payments and refuse corporate bank transfers, accepting only personal accounts or virtual currencies.

真实案例

• In July 2026, Yunnan Net's social channel reported that an AI outbound marketing platform was exposed for carpet-bombing marketing with thousands of calls per day. Number spoofing and de-tagging were standard services, with the platform handling appeals to clean tags. Investigative reporters posing as business owners learned that de-tagging fees were calculated per number.
• In April 2026, Xinhua News Agency published an investigation titled 'Beware of AI Customer Service Phone Scams,' tracking fraud rings using AI to impersonate e-commerce and financial platform support. Calls reached over 1,500 per day, with scripts generated by large models that branched in real-time. Some victims lost tens of thousands of yuan after being led to fake refund links.
• In June 2026, Tencent News reported that the failure of harassment call tagging was exposed. Some AI outbound platforms clean tags by submitting bulk false-positive appeals, allowing tagged numbers to become 'un-tagged' and continue calling. Blocking a single number no longer prevents the same platform from using a different caller ID to continue harassment.
• In April 2026, Xinhua News Agency reported that the Jinchuan Branch of the Jinchang Public Security Bureau in Gansu cracked an AI voice fraud case impersonating official customer service. The gang posed as short-video platform support to induce elderly victims to download apps and perform remote-controlled transfers. 8 suspects were arrested, with over 20 victims and losses exceeding 3 million yuan. (Source: https://www.news.cn/legal/20260428/ccc23e82b9574c7ca1a641d0d10bbf26/c.html)
• In March 2026, NetEase reported on typical cases of unfair competition involving AI, as announced by the State Administration for Market Regulation. The Shanghai Municipal Administration for Market Regulation penalized Shanghai Qiaopin Network Information Technology Co., Ltd., which used AI outbound software to help loan intermediaries impersonate banks to make harassment calls. Authorities ordered the cessation of illegal activities and imposed a fine of 200,000 yuan. (Source: https://m.163.com/dy/article/KPCKRV200511C93M.html)

Official Stance

• On April 28, 2026, Xinhua News Agency published 'Beware of AI Customer Service Phone Scams,' investigating the industry chain and scripts of AI outbound fraud, citing data from anti-fraud departments to warn the public.
• On July 29, 2026, Yunnan Net's social channel released a joint investigation titled 'Thousands of Calls Per Day: The Proliferation of AI Outbound Marketing,' exposing the status of harassment calls and the gray chain of number de-tagging.
• On August 5, 2026, CNR reported 'Over 90% of Fraud Hides in Apps and Network Tools; New Moves in Hardcore Anti-Fraud,' citing National Anti-Fraud Center data showing that over 90% of telecom fraud is implemented via apps and network tools, with AI outbound calls serving as the entry point.

How to Protect Yourself

• ✅ For outbound calls claiming to be from e-commerce, finance, education, or real estate, always hang up and verify through official apps or official website customer service portals; do not follow instructions to add WeChat or click links.
• ✅ Before purchasing telemarketing SaaS, verify the compliance of the number source by requiring suppliers to provide data authorization documents and proof of personal information protection compliance; reject suppliers that offer de-tagging or spoofing services.
• ✅ Enable call warning features from carriers and the National Anti-Fraud Center app on your phone to automatically block tagged harassment calls, and regularly report harassment numbers on the 12321 platform.
• ✅ When signing telemarketing service contracts, insist on corporate account payments and retain records of number sources and script templates as evidence. If gray market characteristics are discovered, report them immediately to local public security cyber-security departments and the MIIT 12321 reporting center.