On January 8, the Commodity Futures Trading Commission (CFTC) Technology Advisory Committee (TAC) held a public meeting. The meeting covered topics including the development of AI, federal cyber defenses, and digital assets.
Below is a summary of the meeting prepared by Delta Strategy Group.
Executive Order on Development of AI
Elizabeth Kelly, Special Assistant to the President for Economic Policy, White House National Economic Council
- AI carries enormous potential benefits and risks. We are very concerned with the potential for AI to undermine human rights while empowering illicit actors. The White House has obtained numerous voluntary commitments from the AI space for transparency and testing to ensure that systems are secure against threats and to reduce consumer threats. While we are somewhat limited in what we can do as the Executive Branch, one of the most important ways to ensure that this innovation happens safely is to lay out guiding principles for anyone in this space to operate on. We are also working to improve competition and release grants for startups and smaller companies to innovate in the AI space. We need to ensure that AI is not used to spread disinformation in a way that could harm consumers or our financial markets. It is vital that development in this space is fully transparent.
Efforts to Modernize federal cyber defenses
Mitch Herckis, Branch Director for Federal Cybersecurity, Office of the Federal CIO, White House
- Both the pace and sophistication of cyber-attacks we are experiencing continue to evolve. It is clear that just strengthening our current defenses will not be sufficient; we will need to create novel and innovative defensive solutions to match the innovation of illicit actors. We believe in a zero-trust approach to cybersecurity, meaning that our defenses are always up for potential attacks. We are aiming for security measures that significantly reduce risk and place the onus on the government, not individuals. There have been significant cultural and technological shifts in the federal government, and we have come extraordinarily far in ensuring sufficient security measures are in place. We have been able to collect vast amounts of data on potential attacks, and we have made significant changes in our system defenses to ensure that we are aligned with the administration’s goals for cyber security.
Understanding the implication of ai on financial markets
Michael Wellman, Professor, University of Michigan
- The speed and precision of AI algorithms in trading are far faster than human reaction times and warrant specific regulatory considerations. They can also scale rapidly, which, again, warrants regulatory attention. It is possible for AI to manipulate markets even if it was not programmed to do so because it is simply looking for the best possible market conditions. There are implications on market performance stemming from AI algorithmic trading, and we must ensure that all market participants are operating on an equal playing field. The key concern is that laws and regulatory processes lag behind technological change. It is also unclear who is responsible for unanticipated uses of technology and who to hold accountable. We believe there are significant concerns with AI being used to manipulate markets to create arbitrage, but it can also be used to create regulatory detectors and prevent manipulation.
Discussion:
Should we be concerned that the Consolidated Audit Trail is already outdated given these AI concerns? Wellman: CAT could be a vital tool, as it contains massive amounts of data that we can train regulatory AI systems on to look for market manipulation.
Subcommittee on Digital Assets and blockchain technology on study of defi
Presentation
- The key issue with DeFi is determining who must be held accountable when things go wrong. People will inevitably try and use the distinction between centralized and decentralized to exploit markets. Regulators must determine whether there are gatekeepers controlling access to markets, as that is a key determination of whether a project is decentralized. It is vital to understand whether there is a governance structure that consolidates control with certain individuals.
- It is apparent that all decentralization exists on a spectrum, so determining how to regulate it is a broad question. It is clear that there will need to be nuance in regulation to account for the differences between different DeFi business structures. This is not just a CFTC issue; this is an issue that all parts of the financial ecosystem must contend with.
- Policymakers must articulate, monitor, and enforce compliance with legal obligations in DeFi. This includes identifying whether the regulatory perimeter might need to be expanded. Regulators also must map counterparty exposure in the DeFi space. All of this must be done through a robust conversation between regulators, academics, and industry.
- This process needs to start with resource assessment, data gathering, and mapping out the industry. Regulators need to identify the extent of the current regulatory perimeter compared with international approaches. The starting point for this should begin with AML considerations.