In August, 2018, The World Economic Forum (“WEF”) published a report, The New Physics of Financial Services, that looks at how AI will transform the financial services sector.

The report's cover photo is of the "Pillars of Creation" from the Eagle Nebula, taken by the Hubble Telescope, which is fitting for a report about how AI will turn the financial sector on its head. The WEF'S analogy is that "AI is changing the physics of financial services...weakening the bonds that have historically held together financial institutions, while creating new centres of gravity where new and old capabilities are being combined in unexpected ways."

The report discusses how matters will change in terms of relationships between participants in the financial industry and addresses a number of issues around the regulation of AI in the financial services sector. A stated goal of the report is to help regulators understand the challenges they face and the necessary responses for both consumers and industry. The observations made and questions raised on that front include:

  • there is no working definition of AI, which leads to confusion and disagreement;
  • data regulators that govern privacy and portability of data will affect the ability of institutions to deploy AI, thus becoming as important as current financial regulators;
  • current regulatory burdens are a heavy strain on financial institutions and AI may help address that through AI that performs regulatory compliance;
  • if regulatory compliance becomes collectively centralized through AI, financial institutions will want to avoid accountability for that centralized performance, creating tension with regulators who want to hold financial institutions accountable;
  • cross-border differences in financial and data regulations will require solutions, but will these solutions be effected?
  • data regulations will affect how institutions enter into partnerships (and data partnerships will be required). Will regulators treat tech companies and financial institutions the same in terms of data-sharing standards?
  • global data regulations are currently in an unprecedented period of change as governments seek to protect citizens. The new rules are not the same and the approaches are not the same across the globe;
  • data regulation will become the critical driver in determining the roles and positioning of players in financial services. Financial services incumbents will be affected by whether or not regulations allow them to access customer data;
  • data sharing complicates regulatory monitoring and auditing. As risk of improper data sharing and breaches rise, regulatory oversight of data management will have to increase;
  • the current regulator talent gap in this field is larger than that experienced by private institutions and this is a challenge to regulators;
  • current regulations may be too complex, uncertain in allocating liability, fragmented, aimed at risk avoidance and require audits (which may be difficult in AI systems). These are roadblocks to implementation of AI;
  • regulators need to address new trends and protect customers but leave room for innovation;
  • "black letter" regulations are needed to ensure clarity so that participants in the financial system know how to comply with regulations. The general principles of consumer financial protection and fairness currently in the regulatory system are sound and do not need to be changed;
  • cross-border coordination is made worse by regional differences in regulation. (this point is very similar to some of the US Treasury report's conclusions).

All of these are good observations. The rest of the report (indeed most of it) refers to the effects of AI on the financial system. It is thoughtful is the result of input from many participants.

This article is revised from a prior summary of the report written by Christopher Alam, who blogs on AI regulation.