The Pulse
Ping An Bank Board Approves First AI Management Rules
Ping An Bank’s board approved internal rules for artificial intelligence on September 29, making it the first listed mainland Chinese bank to publicly disclose such a policy. The rules’ full text remains unpublished as regulators press bank
AI.info Team ·
A board vote, but no public rulebook
Ping An Bank has approved internal artificial-intelligence rules, while keeping the full text from public view. The gap between a formal board decision and an undisclosed policy is the central tension in a move analysts expect other mainland lenders to follow.
The Shenzhen-listed bank’s board approved its AI management measures at a meeting whose voting deadline was September 29, according to a resolution published September 30. All 13 participating directors voted in favour, including five independent directors. South China Morning Post reported that Ping An was the first listed Chinese lender to formally adopt AI rules; Cailian Press said other banks it contacted were implementing the new regulatory guidance and could issue similar policies.
Rules approved as regulator sets expectations
The resolution names the policy, the Ping An Bank Artificial Intelligence Management Measures, but offers no detail about its contents. Cailian Press reported that the bank later described it as its 2026 version 1.0 and said the board’s strategy committee would oversee AI development and use. The bank also described plans for full-lifecycle controls, including data security, algorithm risk, ethics review, responsibility tracing and risk classification.
Those areas track a national regulatory framework issued in June. The National Financial Regulatory Administration’s guidance directs banks and insurers using AI to assign responsibility to a board committee and manage applications from needs assessment and development through deployment, maintenance, evaluation and retirement. It also calls for risk classification, data protection, algorithm screening and clear accountability across business and technology functions.
More than 450 applications already in use
The rules address an expanding body of work inside Ping An Bank. The bank told Cailian Press it had put more than 450 large-model applications into operation by the end of June, spanning targeted marketing, intelligent investment advice, compliance and risk control, and office support.
That scale makes the policy more than a statement about future experimentation: it sets internal controls around systems already used across business functions. The bank said the framework follows the regulator’s June guidance and is intended to support safe, orderly development and use. Yet without the policy’s text, investors, customers and other lenders cannot assess its precise approval thresholds, safeguards or enforcement arrangements.
Peers face the same governance test
Regulation gives banks a common baseline, but leaves institutions to translate it into operating rules suited to their own uses and risk controls. The national guidance sets expectations for committee oversight and lifecycle management, while requiring applications to match a bank’s capacity to manage risk.
“Financial institutions should strengthen their awareness of proactive risk prevention, moving from ‘passive response’ to ‘proactive leadership’ through a combination of institutional design, tool innovation and ecosystem coordination.”
Dong Ximiao, chief economist at Zhaolian and executive director of the Shanghai Finance and Development Laboratory (translated from Chinese)
For now, Ping An’s disclosed step is narrower than a sector-wide shift: one bank has confirmed board approval, and other institutions are preparing their responses. Whether its framework becomes a useful benchmark depends in part on what the bank eventually reveals about the rules it has adopted.