Regulation
Basel Committee to review bank operational-risk loss categories for cyber and AI
After its 28 and 29 September meeting, the global banking standard-setter said frontier AI could amplify operational and cyber vulnerabilities and agreed to test whether its operational-risk loss categories still fit.
On 1 October the Basel Committee on Banking Supervision published the outcome of its meeting in Indonesia on 28 and 29 September 2026. Under the heading Digitalisation, the Committee says it exchanged views on developments in artificial intelligence, and that since its previous meeting on 19 and 20 May 2026 these developments have continued to evolve at a remarkable speed. It adds that the financial footprint of the AI ecosystem is expanding rapidly, accompanied by greater use of leverage and increasingly interconnected financing arrangements.
The release gives a two-sided reading. Frontier AI, it says, offers the potential for substantial opportunities to enhance efficiency and innovation across financial services, but it also has the potential to amplify operational vulnerabilities, including from cyber attacks and correlated dependencies in the financial system. Its integration into critical financial functions, the Committee says, will require careful governance, robust risk management and ongoing supervisory attention. The one concrete step is that the Committee agreed to review the sufficiency and adequacy of the existing event-type loss categories set out in the operational risk framework, with a focus on cyber risk and AI developments, and that it will continue to monitor AI developments and discuss supervisory implications. The release sets no timetable for that review and proposes no new rule.
For people building or using AI trading agents, this is context rather than a requirement. The Basel Committee sets standards for banks, not for retail accounts or software that trades through a broker, so nothing here applies to an individual user directly. What it does show is how the banks that sit behind brokers and market infrastructure are being asked to think about AI: as a source of operational and cyber loss events that their frameworks may need to classify differently, and as a dependency risk when many firms rely on the same systems. This site's coverage of the Bank of England's Financial Policy Committee record points the same way. What to watch next: whether the loss-category review leads to a consultation, and whether national supervisors turn it into expectations for firms that offer agent access to client accounts. The release does not mention trading agents, so any link to them is an inference, not a Committee statement.
Read the original: Bank for International Settlements (Basel Committee) - Basel Committee to review bank operational-risk loss categories for cyber and AI. Commentary is the independent editorial view of Share Trading; the original article is credited to its publisher.