AI news ·
China issues AI governance rules for insurance and banking sectors
China issued new AI governance rules for banks and insurers on June 26, 2026. The move follows data showing 40% of insurance customers now use AI for purchases.
China's National Financial Regulatory Administration released new artificial intelligence governance rules for the insurance and banking sectors on 26 June 2026. The guidelines require financial institutions to ensure their AI systems are compliant, transparent, trustworthy and effectively managed, setting a clear regulatory baseline as AI becomes more deeply embedded in financial services.
The NFRA did not publish the full text publicly, but the announcement signals that regulators want AI models used in underwriting, claims, and customer service to meet strict standards. Institutions will need to demonstrate that their AI tools follow existing financial regulations, that decisions can be explained to customers, and that systems are built with safeguards against bias and error.
What the guidelines require
The four pillars-compliance, transparency, trustworthiness, and effective management-echo themes seen in AI governance frameworks globally. For Chinese financial firms, this means building internal review processes, documenting model logic, and training staff to oversee AI outputs. The rules apply to both proprietary systems and third-party tools, covering everything from credit scoring models to automated claims handlers.
Growing AI adoption in insurance
The regulatory move comes as AI use accelerates across China's insurance industry. A related study found that 40% of insurance customers now turn to AI platforms when deciding on purchases. Insurers are deploying chatbots, automated underwriting engines, and fraud detection models at scale. AI for Insurance is no longer experimental-it's operational, and regulators are stepping in to set boundaries.
Why this matters for insurance professionals
For underwriters, claims managers, and compliance officers, the guidelines mean AI tools can no longer operate as black boxes. Teams will need to document how models reach decisions, test for fairness, and establish escalation paths when AI errors occur. The rules also signal that regulators may audit AI systems during routine examinations. Insurance professionals who understand both the technology and the regulatory expectations will be better positioned as their firms adapt. For banking counterparts, similar expectations apply, as AI for Finance now faces formal regulatory scrutiny.