Today, Congresswoman Maxine Waters (D-CA), the top Democrat on the House Financial Services Committee, sent a letter to Congressman Patrick McHenry (R-NC), Chair of the Committee, urging McHenry to immediately convene a hearing to explore the implications of generative Artificial Intelligence (AI) on housing and financial services. In the letter, Ranking Member Waters sounds the alarm on the rapid growth of AI technologies, and the leading role the Committee must play in ensuring that these technologies do not exacerbate existing disparities and inequities within our financial system.

In 2019, one of the first moves then Chairwoman Waters made after assuming the gavel was to establish Congress’ first-ever Task Forces on Financial Technology and Artificial Intelligence. The historic Task Forces held a total of 22 hearings across the 116th and 117th Congress and led the way to encourage equitable and responsible innovation. The Task Forces also worked to ensure that regulators and the law were adapting to the changing landscape to best protect consumers, investors, and small businesses.

“I write to request that you immediately schedule a Committee hearing on the implications of ChatGPT, Microsoft Bing’s Artificial Intelligence (“AI”) chat, and other AI advancements for the housing and financial services industry,” said Ranking Member Waters. “With the rapid growth of AI technologies, including Machine Learning (“ML”), companies and industry stakeholders can process large amounts of data and make complex consumer and investment decisions with minimal human direction. Recent reporting has revealed that financial institutions are using and testing AI in their businesses, as well as hiring AI talent. From tenant screening to home appraisals, the housing industry is also rapidly developing its use of AI technologies to make critical decisions about who has access to housing and the value of someone’s largest investment—their home—often with discriminatory adverse effects. Federal agencies are also taking steps to oversee the use of AI, including in home appraisals, with a recent joint proposed rule from six agencies overseeing the financial industry meant to protect consumers from inaccurate or discriminatory AI decisions. If harnessed correctly, AI can expand financial access for consumers, aspiring homeowners and renters, and businesses. However, the speed at which AI is being developed can potentially outpace Congress’ and regulators’ roles in fully assessing and understanding the harms of AI systems and the ability to put in place integral consumer and investor protections.”

Read the full letter here and below.

The Honorable Patrick McHenry
Chairman
House Committee on Financial Services
2129 Rayburn House Office Building
Washington, DC 20515

Dear Chairman McHenry:

I write to request that you immediately schedule a Committee hearing on the implications of ChatGPT, Microsoft Bing’s Artificial Intelligence (“AI”) chat, and other AI advancements for the housing and financial services industry. With the rapid growth of AI technologies, including Machine Learning (“ML”), companies and industry stakeholders can process large amounts of data and make complex consumer and investment decisions with minimal human direction. Recent reporting has revealed that financial institutions are using and testing AI in their businesses, as well as hiring AI talent.^[1] From tenant screening to home appraisals, the housing industry is also rapidly developing its use of AI technologies to make critical decisions about who has access to housing and the value of someone’s largest investment—their home—often with discriminatory adverse effects.^[2] Federal agencies are also taking steps to oversee the use of AI, including in home appraisals, with a recent joint proposed rule from six agencies overseeing the financial industry meant to protect consumers from inaccurate or discriminatory AI decisions.^[3] If harnessed correctly, AI can expand financial access for consumers, aspiring homeowners and renters, and businesses. However, the speed at which AI is being developed can potentially outpace Congress’ and regulators’ roles in fully assessing and understanding the harms of AI systems and the ability to put in place integral consumer and investor protections.

On November 30, 2022, OpenAI, a research company led by Sam Altman and backed by Microsoft, debuted ChatGPT, a generative AI language model. The chatbot quickly went viral, attracting one million users within five days.^[4] Various versions of ChatGPT have been released with each version exponentially building upon previous models. The most recent version, ChatGPT-4, was announced on March 13, 2023. As market competitors such as Google’s Bard AI and DeepMind’s Chinchilla AI continue to examine their generative AI projects, the technology’s capabilities continue to expand.^[5] The Committee must move quickly to consider the potentially disruptive effects of these powerful AI models.

As Chair of the Committee, I responded to the emergence of new technologies in the housing and financial services industry by establishing Congress’ first-ever Task Forces on Financial Technology (“Fintech”) and Artificial Intelligence. Under my leadership, Committee Democrats encouraged equitable and responsible innovation and sought to ensure that regulators and the law were adapting to the changing landscape to best protect consumers, investors, and small businesses.

For four years, the Task Force on Artificial Intelligence (“AI Task Force”), held numerous hearings on the policy implications of the use of AI technologies, including how AI can be used to assess risks, predict outcomes, and allocate capital across the financial system in a more efficient manner than traditional human assessments. However, the AI Task Force also raised ethical and legal concerns related to algorithmic bias and discrimination, particularly if automated programs do not perform as intended or have adverse effects on members of protected classes. These technologies may embed historic inequities in the financial services and housing markets through the use of data that reflect underlying bias or discrimination. Beyond ethical risks, large-scale societal and economic harm is also possible if AI’s growth implications are not monitored closely. Generative AI also poses dangers in the financial sector such as consumer and employee data exposure, providing inaccurate financial assessments, and spreading misinformation that creates a broader risk to financial stability. Further, we often lack transparency in the training data, challenging our trust in the decision-making of these models. In recent months we have seen the exponential growth of generative AI models and must continue to do our part in examining these issues. On Tuesday, May 16, Speaker Kevin McCarthy and Minority Leader Hakeem Jeffries convened a bipartisan briefing with the CEO of OpenAI and creator of ChatGPT, Sam Altman. Later that day, Sam Altman, testified before the Senate Judiciary Subcommittee on Privacy, Technology, and the Law, stating that government intervention will be necessary to mitigate the risks of increasingly powerful AI systems.

We have seen the Biden Administration take meaningful steps in promoting responsible innovation while protecting Americans. This encompasses the landmark Blueprint for an AI Bill of Rights^[6] including related executive actions,^[7] the AI Risk Management Framework,^[8] and a roadmap for standing up a National AI Research Resource.^[9] Our Committee must do its part by examining the issues raised in the financial services sector.

I hope we can continue working together on these issues and respectfully request that you schedule a Committee hearing on the implications of these powerful AI systems in housing and financial services without delay.



Sincerely,

MAXINE WATERS
Ranking Member
Committee on Financial Services


______________________________________
[1] William Shaw & Aisha S. Gani, Wall Street Banks Are Using AI to Rewire the World of Finance, Bloomberg (May 31, 2023).
[2] Khari Johnson, Algorithms Allegedly Penalized Black Renters. The US Government Is Watching, Wired (Jan. 16, 2023); See also Urban Institute, How Automated Valuation Models Can Disproportionately Affect Majority Black Neighborhoods (Dec. 2020).
[3] Quality Control Standards for Automated Valuation Models (June 1, 2023). Agencies involved are the Consumer Financial Protection Bureau, the Board of Governors of the Federal Reserve System, Federal Deposit Insurance Corporation, Federal Housing Finance Agency, National Credit Union Administration, and the Office of the Comptroller of the Currency.
[4] A Short History Of ChatGPT: How We Got To Where We Are Today, Forbes (May 19, 2023).
[5] Top Large Language Models (LLMs) in 2023 from OpenAI, Google AI, Deepmind, Anthropic, Baidu, Huawei, Meta AI, AI21 Labs, LG AI Research and NVIDIA, MarketTechPost (Feb. 22, 2023).
[6] The White House, Blueprint for AI Bill of Rights (accessed on May 19, 2023).
[7] The White House, Executive Actions related to Artificial Intelligence (Oct. 4, 2022).
[8] The White House, AI Risk Management Framework (accessed on May 19, 2023).
[9] The White House, Roadmap for Standing up a National Research Resource (Jan. 2023).

###