State Lawmakers Wrestle with Predictive Pricing

The days of simply setting a price for goods or services and waiting for customers to accept it are gone. Thanks to predictive pricing, businesses of all sorts can now use data analytics, machine learning and algorithms to anticipate market demand and adjust prices in real time, a la the airline industry.

The technology is a boon to businesses, as it allows them to optimize their prices and maximize profits. Recent actions by the Federal Trade Commission and the U.S. Department of Justice, however, have highlighted its potential for abuse, leading state lawmakers across the country to seek to regulate its use, with an eye towards defining what conduct is legal and what isn’t.

“It doesn’t matter if price fixing happens behind closed doors or through artificial intelligence, its wrong either way,” said California Assemblywoman Cecilia Aguiar-Curry (D) in a January press release announcing the introduction of AB 325, which would allow the state attorney general to sue companies that use price algorithms trained on “nonpublic competitor data.”

AB 325 is indicative of the kind of nuanced issues legislators are wrestling with when it comes to predictive pricing, otherwise known as dynamic pricing. At its core, predictive pricing is simply using artificial intelligence to adjust prices in response to market conditions. There’s nothing inherently wrong with that.

But how predictive pricing strategies assess market conditions and respond is potentially a problem and where legislators are looking to make a difference. Aguiar-Curry’s bill addresses the issue by targeting nonpublic competitor data, with the thinking apparently being that predictive pricing is similar to an illegal agreement among competitors to set prices and undercut competition, i.e. price fixing.

“AB 325 shows California is working to protect our people and ensuring a competitive marketplace for all so AI isn’t misused to jack-up prices,” the assemblywoman said in her press release.

California Leading Way

AB 325 is just one of at least five bills that have been introduced in the Golden State to address AI’s potential to harm consumers. The others are:

• AB 446, which would ban businesses from using the personal information of consumers to change the price of goods or services.

• SB 52, which would prohibit “algorithmic devices” used to set prices or occupancy rates for rental properties.

• SB 259, which would bar “affinity-based algorithmic pricing” that sets prices using “perceived characteristics from personal data.”

• SB 295, which would require companies with revenues of more than $5 million annually to make detailed reports on the pricing algorithms they use, including their data sources and any reliance on competitor information.

“Imagine a shady character like Bob whispering pricing secrets to businesses, urging them to inflate costs to maximize profits,” SB 295’s sponsor, Sen. Melissa Hurtado (D), said in a press release. “Bob would be behind bars, so why should we tolerate algorithms doing the same thing in the shadows?”

Over Dozen States Targeting Predictive Pricing

So far this year at least 14 states have introduced bills dealing with predictive pricing—dynamically adjusting prices for products and services using artificial intelligence to maximize profits—according to the State Net® legislative tracking system.

Other States Considering Issue

California isn't the only state where legislators are looking at this issue. According to LexisNexis® State Net® data, at least 26 bills dealing with predictive pricing have been introduced in 14 states since the start of the year.

Several of those bills seek to address algorithmic price fixing of rental properties and housing: Arizona’s HB 2847; Connecticut’s HB 6947; Hawaii’s HB 831 and its companion measure SB 157; New Mexico’s HB 215; and New York’s AB 1417 and AB 3930.

Idaho’s HB 203 would prohibit service providers from using a pricing algorithm that “incorporates nonpublic data from two...or more sellers or buyers of a similar good or service.” Ohio’s SB 79 would make the use of a pricing algorithm that “uses, incorporates, or is trained with nonpublic competitor data a violation” of the Buckeye State’s antitrust law. New York’s AB 4427 would bar insurance rates from being determined using algorithms or predictive models leveraging external consumer data.

Conflicting Legal Rulings

In the courts, meanwhile, there have been divergent rulings on whether software can violate competition laws. In May, a Nevada federal judge rejected the notion of price-fixing through the use of algorithmic tools when she ruled in favor of Las Vegas strip hotels that use such software to set their room rates.

But then in December a Washington state federal judge denied a similar motion to dismiss another suit involving algorithmic pricing antitrust claims.

As these cases illustrate, the issues involved here are nuanced, a point California Assemblywoman Aguiar-Curry makes in a fact sheet about her bill to penalize companies that use price algorithms trained on nonpublic, competitor data.

“Price fixing algorithms are designed to artificially inflate prices and restrict supply and undermine fair market competition. Price fixing is illegal under existing law, but price fixing by algorithms is challenging to detect,” states the fact sheet on AB 325. “This problem is compounded by problematic legal precedent that requires additional evidence of a price fixing scheme before that information is known to enforcers.”

In a statement to SNCJ, the assemblywoman said, “AI price fixing is anti-competitive, restricts supply, and artificially inflates prices, and most Californians don’t even know this is happening.”

—By SNCJ Correspondent BRIAN JOSEPH

Visit our webpage to connect with a LexisNexis® State Net® representative and learn how the State Net legislative and regulatory tracking service can help you identify, track, analyze and report on relevant legislative and regulatory developments.