ICBC President Liu Jun: Financial institutions urgently need to improve their ability to accurately identify and price non-economic risks.

How can AI and intelligent agent-based financial services mitigate multidimensional risks?

On March 23, Liu Jun, president of the Industrial and Commercial Bank of China, stated at the China Development Forum 2026 Annual Conference that the current uncertainty in the international economy is increasingly dominated by non-economic fields. Global risks are no longer a linear time series; they have evolved into a multidimensional risk matrix.

Liu Jun believes that in the face of rapidly evolving risks, it is necessary to reshape a credible international cooperation system, which can be called “Globalization 2.0.” This means that by jointly addressing global issues such as climate change and AI governance, we can inject certainty into a transforming world to reduce risk premiums and achieve win-win outcomes for all parties. For financial institutions, this transformation requires a fundamental shift across three strategic dimensions.

First, reprice non-economic and non-market risks. Liu Jun pointed out that in the past, financial institutions primarily focused on credit risk, market risk, liquidity risk, and so on. Today, the variables they face are more complex and dynamically evolving. Risks such as geopolitical tensions and wars can develop into systemic risks. Traditional risk models based on historical data and empirical rules are now difficult to sustain. Therefore, there is an urgent need to enhance the ability to accurately capture and price non-economic risks, utilizing big data, AI, remote sensing technology, and other tools to construct an engineered financial risk management system for the scientific quantification of various extreme risks.

Second, cultivate π-type talents for the digital age. Liu Jun believes that in the AI era, innovation increasingly relies on deep vertical research and understanding in specific fields. “We need π-type talents because they possess both horizontal career expansion and vertical professional depth. The ideal π-type talent should not only have two pillars of expertise but also possess the ability for deep research across multiple fields to better address the challenges posed by AI and the integration of knowledge brought by intelligent agents.”

Third, move towards a multidimensional, systematic, and intelligent agent-based financial service model. Traditional flat and passive financial services can no longer meet the new demands arising from the restructuring of global patterns and rising economic uncertainty. Therefore, it is necessary to deeply integrate support across the entire lifecycle and services across the entire industry chain, building a systematic financial service framework for the real economy. Such financial institutions will transform into comprehensive service providers of capital, information, and efficiency, allowing for the buffering and resolution of risks at individual nodes within a multidimensional grid.