How can financial institutions adopt generative AI responsibly?

Firms need to weigh risks and benefits of new technology as they deploy it within their organizations.

In brief

• Trust and accuracy of the output created by generative AI remain key areas of concern.
• As governments deliberate potential action, financial institutions need to adopt frameworks that assess and manage risks.
• A phased approach, with experimentation and scaling of use case implementation, to support a responsible AI activation framework is essential.

While generative AI holds tremendous potential, it is also constrained by the risks and limitations associated with this technology. Concerns have been raised related to what might occur from improper use of these technologies, absent adequate guardrails. There is also a growing apprehension about how these models will disrupt the workforce. Furthermore, the financial and sustainability implications of using powerful large language models have yet to be addressed.

To address this, some industry observers are urging national and local governments to accelerate AI regulation in concurrence with its adoption. The European Parliament has already taken preliminary action, passing a law that would attempt to regulate artificial intelligence. In the US, the White House published an AI bill of rights last fall, while the US Senate unveiled its plan to discuss potential AI regulations.

While these regulations have yet to take effect, they could foreshadow actions governments and other regulatory bodies might take to establish guardrails for this rapidly developing technology.

Until a consensus is reached on what those guardrails should be, financial institutions must carefully review the risks in context of the intended use and set up appropriate evaluation and risk mitigants to limit the exposure from activating these technologies in their nascent stage.

Certain key risks and considerations are highlighted below: