Housing and high finance are the two sectors of the economy where government regulation of artificial intelligence (AI) is most needed, a House Financial Services Committee working group has concluded after a series of meetings with industry players.
The working group said the committee should “play a leading role in overseeing the adoption of A.I. in the financial services and housing industries” in a new report.
The report is from the House Financial Services’ 12-member AI working group, which launched in January and includes both Democrats and Republicans. The lead authors of the report are committee Chair Patrick McHenry (R-N.C.) and ranking member Maxine Waters (D-Calif.).
It paints a picture of AI as becoming infused within numerous financial subsectors and stitched into the transactional fabric of the economy.
It describes how AI is being used in capital markets for widespread surveillance, raising questions about the extent to which conventional forces are even still operational in securities markets in which algorithmic momentum has been driving as much as 70-percent of daily trading, according to market researchers and former Federal Reserve bankers.
Brokers and investors told the committee that AI is reducing price volatility levels through channels such as the timing of trades.
The report also details how AI is proliferating across business practices including loan underwriting, fraud detection and debt collection, as well as customer service more broadly.
Questions about privacy and the amount of information that companies can acquire about customers through the use of AI run through the working group’s analysis.
In one instance, businesses told the committee about using “computer vision technology to verify know-your-customer (KYC) information” as a way to fight fraud.
In the housing sector, different kinds of businesses are using AI to underwrite mortgages and insurance policies, screen tenants and do data analysis.
The committee said these uses present challenges to fair housing practices and consumer protection as well as anti-discrimination laws.
“AI could lead to bias and discrimination and make it harder to detect such outcomes due to a lack of explainability,” it noted.
Regulators who participated in the discussion observed that businesses would be expected to “follow all laws, including anti-discrimination and other consumer protection laws, in a tech-neutral manner.”
AI and algorithms more broadly have come under fire recently in various sectors of the economy within the context of firms’ pricing strategies. Authorities and researchers say they can result in forms of price fixing that can effectively turn competitive industries into collusive cartels.
The Department of Justice recently got involved in one New Jersey lawsuit involving the hospitality sector, filing a statement of interest in a “hotel room algorithmic price-fixing case.”
“Companies across the economy are increasingly using algorithms to determine their prices. When a small group of algorithm providers can influence a major segment of a market, competitors are better able to use the algorithm provider to facilitate collusion,” the DOJ said in a March statement.
“Competitors cannot lawfully cooperate to set their prices, whether via their staff or an algorithm, even if the competitors never communicate with each other directly,” the agency warned.
Academic research confirms the cartelizing effects of algorithmic pricing.
“Widespread adoption of the same algorithm could also lead to price coordination, resulting in elevated prices,” researchers at the University of Pennsylvania concluded in a 2023 study of software used in the multifamily housing rental market.
They found that more algorithmically attuned markets had “higher rents and lower occupancy” and that the patterns they observed were “consistent with either price coordination through the algorithm or widespread pricing error.”
Algorithmic price fixing has been on the radar of investigative journalists and lawmakers for the past few years.
ProPublica investigators found in 2022 that rents for 70 percent of all apartments and houses in one Seattle neighborhood, which were operated by ten different property managers, were all set using pricing software for a single company called Real Page, raising questions about the extent to which “markets” are even functioning in algorithmic price environments.
“RealPage not only provides clients access to anonymized, non-public information, but has created a forum – RealPage User Group – which encourages landlords to work together on issues like revenue management,” Sens. Elizabeth Warren (D-Mass.), Tina Smith (D-Minn.), Edward Markey (D-Mass.) and Bernie Sanders (I-Vt.) wrote in a letter to the company in 2022.
A different pricing behavior enabled by high-powered algorithms known as surge pricing or dynamic pricing, in which prices change moment to moment, was also addressed in the report released on Thursday.
One housing sector professional told the committee that dynamic pricing needed oversight, noting “potential for collaboration by market participants in dynamic pricing algorithms.”
The person said federal agencies should look into the use of rent-setting technologies.
Calls to regulate AI are sounding forth internationally, as well.
Technology advocates in the United Kingdom, where a new Labour government has just taken power with a sweeping reform agenda after 14 years of Conservative rule, are saying they want to see more done on AI.
“What worries me in terms of the way that this government is approaching AI is that they’re not being inclusive and talking to the rest of civil society, as well as top academics and researchers,” technology commentator Stephanie Hare told the BBC Wednesday.
Updated at 2:06 p.m. EDT