On July 31, the National Financial Regulatory Administration and the Cyberspace Administration officially released the "Guiding Opinions on Promoting the Application of Artificial Intelligence in Financial Risk Control" (hereinafter referred to as the "Opinions"), proposing to launch regulatory sandbox pilot programs in some cities, allowing financial institutions for the first time to incorporate blockchain-based on-chain credit data into credit approval and risk control models. This marks the first time that domestic regulators have provided an institutional channel for the application scenario of "on-chain data + AI risk control", injecting new vitality into the transition of DeFAI (Decentralized AI Finance) from concept to implementation.
Policy Implementation: On-chain Credit Data Receives Official Endorsement
According to the "Opinions", the regulatory sandbox will first be implemented in licensed financial institutions in Beijing, Shanghai, Shenzhen and other places. Pilot institutions can apply to use on-chain data such as transaction records, lending records, and asset proofs recorded through blockchain as a supplement to traditional credit reporting. At the same time, regulators require AI risk control models to have interpretability and auditability to avoid algorithmic discrimination and systemic risks.
The industry generally believes that this means that on-chain credit assessment, which has long been in a gray area, has received an "official pass". Previously, due to the lack of clear rules, financial institutions were very cautious in using on-chain data, while technology companies with data accumulation and technical reserves found it difficult to enter the compliant market. This policy breakthrough is expected to drive investment opportunities in the complete industry chain from data provision, model training to underlying computing power.
AI Risk Control Concept Stocks Rally Sharply
After the policy announcement, related concept stocks in Shanghai, Shenzhen, Hong Kong and US markets all showed rapid increases. In A-shares, the digital finance sector, known for "intelligent credit", led the gains, with some individual stocks opening up more than 7%; in Hong Kong, it manifested as a joint rise in internet technology stocks and fintech stocks. Notably, "blockchain + AI" stocks, which had been weak due to compliance issues, also showed a clear rebound.
In overseas markets, AI lending leader Upstart Holdings (UPST) and SoFi Technologies (SOFI), which have already布局加密信贷业务, both saw pre-market gains between 3% and 5%. Market analysts pointed out that although these companies mainly focus on traditional financial scenarios, they have all begun to explore the integration points of on-chain data and AI risk control, and the policy benefits have directly enhanced their future revenue expectations.
DeFAI Concept Stocks Become the Strongest Theme
Among the many AI concept stocks, the DeFAI sector performed most prominently. The so-called "DeFAI", or Decentralized AI Finance, refers to using AI technology to optimize risk pricing, liquidation, and yield aggregation in DeFi protocols. Due to the natural characteristics of on-chain data being transparent, tamper-proof, and traceable, it is particularly suitable as a basis for training and validating AI models. Therefore, the opening of the regulatory sandbox is seen as a prerequisite for DeFAI to enter the mainstream.
Some analysts believe that the regulatory authorities' choice of on-chain credit data as a breakthrough point is the result of careful consideration. On the one hand, traditional credit scoring systems have been highly mature in China, and on-chain data can serve as alternative data for "long-tail populations", enhancing the coverage of inclusive finance. On the other hand, through controllable sandbox pilots, the real performance of AI models in scenarios such as cross-border payments and encrypted asset collateral can be observed, accumulating experience for subsequent comprehensive legislation.
Investment Logic: Grasping the Three Elements of "Data + Algorithm + Scenario"
From an investment perspective, the current rally of AI risk control concept stocks is not simply event-driven, but has a mid-term industrial logic. First is the "data layer", companies with the ability to legally collect and govern on-chain data will become scarce resources; second is the "algorithm layer", fintech companies with mature AI risk control models and interpretability technology will benefit preferentially; finally is the "scenario layer", platforms that can use on-chain credit for specific businesses such as consumer finance, micro-loans, and encrypted collateral lending will open new revenue curves.
However, investors should also pay attention to risks. Although the regulatory sandbox provides experimental space, the final commercialization still has uncertainties, especially regarding data privacy protection and model fairness issues, which may require multiple iterations. The sector has seen large short-term gains, and there is a possibility of emotional pullbacks. For long-term capital, screening companies with real business scenarios and technical barriers is more critical than chasing concepts.
Conclusion
The formal entry of on-chain credit data into the regulatory sandbox is a milestone in the integration process of AI risk control and DeFAI. It marks that supervision is shifting from "defensive supervision" to "proactive guidance supervision", and the industry is expected to achieve explosive growth within a compliant framework. For investors, understanding the industry trends behind the policy is far more important than short-term K-lines.
