Token Trading and Fraud Disrupt China's AI Service Market
Fraudsters and underground traders are exploiting AI service tokens in China's AI market through off-platform schemes, fake investment products, and mass account farming, causing economic losses to platforms and users. Regulators and legal experts say weaknesses in legal definitions, cross-agency coordination, and platform enforcement allow speculation and fraud to proliferate.
What happened
Token platforms intended tokens as non-transferable service units for individual users only. However, off-platform trading emerged through WeChat groups, QQ groups, second-hand platforms and forums, creating an underground market with bulk hoarding, reselling and price arbitrage.
Fraudsters repackaged tokens as 'high-appreciation investment products' and 'guaranteed return wealth management products' using slogans like 'hoard tokens to get rich quick' and 'AI digital gold,' enticing public speculation. Some non-licensed operators created fake 'token price indices' and operated leverage trading and contract trading platforms, while others structured token-like schemes as 'token agent' multi-level distribution networks resembling pyramid schemes.
Bad actors exploited platform account registration and subsidy loopholes by automatically registering thousands of fake accounts, claiming free tokens and platform subsidies, then selling stolen tokens at low prices for profit. Token daily call volume grew from 100 billion in early 2024 to 1 trillion by end of 2025 and 1.4 trillion by March 2026.
Reported impact
Fraudulent activities inflicted direct economic losses on platforms and raised costs for legitimate users and small developers, eroding innovation returns and disrupting market competition.
Off-platform trading at premium prices above official rates distorted platform pricing mechanisms and destabilized regular supply systems, particularly for scarce scenarios like high-precision generation and long-context tasks.
Index analysis
The expansion of token speculation reflects multiple overlapping factors: legal uncertainty about whether tokens constitute data products, virtual property, or functional rights; gaps between rapidly evolving AI practices and existing competition, anti-fraud and securities regulations; and limited cross-agency coordination among internet, market supervision, financial, police and industry authorities.
Platform control gaps include weak real-name authentication enforcement allowing batch fake account registration, incomplete transaction record-keeping and traceability systems unable to detect abnormal hoarding or mass scraping, and insufficient monitoring or weak penalties for off-platform trading and resale.
Governance controls
Officials cited the absence of clear legal definition for tokens and noted that monitoring departments faced confusion about jurisdiction, legal authority, and behavioral boundaries, creating gaps where speculation could proliferate without enforcement.
Authorities recommended establishing legal frameworks defining tokens as metering units for AI services with no financial attributes, unified data element transaction rules, cross-agency risk monitoring mechanisms, platform responsibility enforcement including real-name verification and transaction traceability, and industry self-regulation through guild agreements.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.