Yellen made the remarks during a keynote address at a conference on AI and financial stability on June 6. Her speech stressed the urgency of addressing these emerging risks and called of both the government and private sectors to collaborate on finding solutions.

The Treasury secretary said:

“This [AI] is a rapidly evolving field. We have our work cut out for us.”

Increasing risk

Yellen acknowledged the advancements AI has brought to the financial sector, such as improved fraud detection and enhanced customer service through chatbots.

However, she and other experts cautioned that deeper integration of AI could lead to increased risks, including the potential for AI to be misused in scams or market manipulation through misinformation.

Yellen warned:

“Insufficient or faulty data could also perpetuate or introduce new biases in financial decision-making.”

She highlighted the complexities of AI models, the inadequacies in current risk management frameworks, and the reliance on a limited number of models by numerous market participants as key areas of concern.

The Treasury Department has issued a request for information to gather insights from stakeholders about the uses, opportunities, and risks of AI in the financial services sector. This initiative is intended to inform future policymaking by incorporating expert opinions and current practices.

Yellen said:

“The tremendous opportunities and significant risks associated with the use of AI by financial companies have moved this issue toward the top of Treasury’s and the Financial Stability Oversight Council’s agendas.”

AI under scrutiny

Yellen’s warning comes amid a wider governmental scrutiny of both AI and the companies behind the technology. The DOJ is reportedly preparing to investigate multiple tech giants, including Nvidia and Microsoft, over antitrust and competition concerns related to AI technology.

US antitrust enforcer Jonathan Kanter announced plans to investigate the AI sector due to concerns about potential monopolies, according to a Financial Times report.

Kanter highlighted the need to examine AI’s competitive landscape, focusing on areas like computing power, data for training large language models (LLMs), cloud services, engineering talent, and hardware.

Kanter emphasized the urgency of acting to prevent dominant tech firms from monopolizing the AI market. He suggested real-time regulatory intervention to be effective and less invasive.

A particular concern is the scarcity of graphics processing units (GPUs) necessary for training LLMs, with rising demand impacting chip allocation.

Mentioned in this article