Financial firms shift from model perfection to systematic engineering

Yu Feng, chief information officer of Guotai Haitong Securities, mandates a systematic engineering approach to AI governance, arguing that the financial industry cannot wait for a perfect model (CF40 Research Insights on China’s…). He directs his team to build a chain of responsibility that spans the sources of knowledge, the creation of records, and the functions AI is authorized to perform. This framework confines AI services to a scope that can be monitored, rolled back, and constrained, ensuring humans remain responsible for setting the direction and bearing accountability for every decision.

Industrial and Commercial Bank of China president Liu Jun describes this as solving an identity dilemma, where firms struggle to assign legal responsibility among developers, deployers, and end users. World Bank survey data confirms the urgency of this transition, noting that only 11% of financial regulators in emerging markets currently use AI supervisory tools, while 50% report a lack of the personnel and tools needed for effective oversight. Zhang Wencai of the World Bank proposes that these institutions strengthen their foundations by incorporating digital identity into payment systems and ensuring financial institutions remain accountable for AI-driven decisions regardless of the technology's origin.

• Nigeria’s Data Protection Commission now provides a legal framework for the lawful and accountable processing of personal data to address AI governance challenges (Global Voices).
• France has published the first part of a public-sector AI procurement guide to help buyers assess AI solutions before contractual deployment (Ctrl+AI+Reg).