
大型银行监管机构应监督银行如何使用人工智能,但不应将其视为与其他风险不同的独立风险——这是周三发布的新监管指引所提出的要求。当前,全球监管机构正加大力度遏制AI驱动的风险。
Big bank supervisors should monitor how lenders use artificial intelligence but should not treat it as distinct from other risks, according to new regulatory guidance issued on Wednesday, as authorities globally ramp up their efforts to contain AI-driven risks.
美国货币监理署负责监管大型全国性银行。据路透社看到的一份文件副本,该机构发布了其称为史上首份关于检查人员应如何审查银行AI使用情况的指引。
The Office of the Comptroller of the Currency, which oversees large national banks, issued what it said was the first-ever guidance for how examiners should probe banks' AI usage, according to a copy of the document seen by Reuters.
根据路透社率先报道的这一新框架,货币监理署表示,银行无需为AI使用设立单独的风险参数,检查人员也不应将这项技术视为一种独特的风险。但检查人员应考虑银行使用AI的方式,以及银行的规模和复杂程度。
Under the new framework, first reported by Reuters, the OCC says banks are not expected to establish separate risk parameters for AI use and examiners should not treat the technology as a unique risk. But examiners should consider the way a lender is using AI, as well as the size and complexity of the lender.
尽管该文件不像法规那样具有法律约束力,但银行通常会严格遵守此类指引。
While the document is not legally binding like a rule, banks typically comply closely with such guidance.
全球各地的银行监管机构一直在向银行询问其AI使用情况,担心这项技术构成重大的网络安全风险,并可能带来金融稳定风险。路透社此前报道,在美国,监管机构一直在询问银行使用先进AI模型的情况、银行对这些系统施加了哪些限制以及如何对其进行监控。
Banking regulators worldwide have been querying lenders about their AI use, worried the technology poses major cybersecurity and potentially financial stability risks. In the US, regulators have been asking banks about their use of advanced AI models, what limits banks place on those systems and how they monitor them, Reuters previously reported.
货币监理署表示,其检查人员应关注银行因AI可能出现的失误所造成的后果,而不是银行所用特定模型的复杂程度。
The OCC said its examiners should focus on the consequences of potential AI-driven mistakes at banks rather than the sophistication of particular models they're using.
例如,该机构表示,一个被授权发起交易的简单AI代理,可能被认为比一个用于内部目的的复杂模型风险更高。文件称,风险较高的AI用途可能需要对自主性施加更大限制,或需要更直接的人工监督。
For example, it said a simple AI agent that is authorized to initiate transactions could be considered riskier than a sophisticated model deployed for internal purposes. Higher-risk uses of AI may require greater restrictions on autonomy or more direct human oversight, the document said.