CFTC Commissioner Sets Out Regulatory Approach Over AI in Derivatives Markets
Commentary by Steven Lofchie
CFTC Commissioner Kristin N. Johnson described a three part approach to regulating AI in futures and derivatives markets.
At an FIA Law & Compliance Division Conference on Innovation in CFTC regulated markets, Ms. Johnson urged the CFTC "to consider three specific interventions:" (i) heightened penalties when AI is used to engage in fraud or market manipulation; (ii) the adoption of "a principles-based regulatory framework for addressing the increasing prevalence of AI in [derivatives] markets;" and (iii) "an inter-agency task force to consider the adoption of parallel, harmonized safeguards that will focus on ensuring the stability and integrity of our markets." (See related coverage.)
Given that the market is already deploying AI based tools, Ms. Johnson suggested that the CFTC should ensure that existing AI regulation and the development of new rules (i) are "fit for it's purpose," (ii) effectively address risks and (iii) align with existing compliance obligations. Ms. Johnson said that "existing governance requirements can also help facilitate effective oversight of AI applications."
Ms. Johnson stated that a new CFTC principles-based framework should address: (i) promoting the explainability of AI models; (ii) the need for data controls; (iii) implementing measures to address bias; (iv) a focus on the governance of AI models; and (v) testing and monitoring output. Ms. Johnson referred to the current work on the framework under her sponsorship of the CFTC Market Risk Advisory Committee including (i) conducting a survey of registrant's use of AI; and (ii) potential recommended guidance or rulemaking "based on how CFTC market participants are using AI," and "any gaps identified in existing regulations and guidance." (See related coverage.)
As to the creation of a new task force, Ms. Johnson said it should focus on information sharing and develop "guidelines, tools, benchmarks, and best practices for the use and regulation of AI in the financial services industry."
Commentary
Commissioner Johnson says that there should be "heightened penalties" for the use of AI to engage in fraud. Why? Shouldn't the penalties be reflective of the fraud and not of the use of AI? It seems absurd to argue that penalties should be set at a lower level for fraud that does not use AI, but that is the flip side of Commissioner Johnson's argument. Rather than treating AI as an overwhelming threat to the financial system that is fundamentally different from all else (fraud is bad, and the use or non-use of AI in the fraud does not change the nature of the violation), the regulators should take a use-specific approach to regulation. (See also FIA Cautions CFTC on Regulation of AI.)
Related Articles
-
CFTC Commissioner Kristin N. Johnson reiterated her support for regulatory actions that foster responsible use of AI in financial markets.
-
At the first Market Risk Advisory Committee of the year, CFTC Commissioner Kristin N. Johnson outlined the latest findings and recommendations from the MRAC Subcommittees on (i) Central Counterparty Risk, (ii) Market Structure and (iii) Climate-Related Risk.
-
CFTC Commissioner Kristin N. Johnson warned against the "potential perils of integrating innovation without proper guardrails."
