|
Who Will Regulate AI?
The reconciliation package working its way through Congress — dubbed the One Big Beautiful Bill Act — could unwind state-level efforts to regulate artificial intelligence.
Buried within the legislation is a provision that would prohibit states from regulating AI and automated decision technology over the next decade. As of Thursday, the bill had passed in the House.
So far, regulating AI has largely been left to the states. Altered images, audio and media are addressed through legislation passed in Alabama and Utah, according to Bryan Cave Leighton Paisner. California has several AI-related laws on the books. Among them, developers in the Golden State must disclose whether an individual’s data is used to train AI models, and if content or advertising was created using generative AI, among other protections. Several states give consumers the right to opt out of companies using their data for automated decisions.
More recently, state legislators have emphasized transparency — disclosing how models are trained, what information is used and when consumers are interacting with a bot, says Goli Mahdavi, a partner and global AI lead at Bryan Cave. State lawmakers are concerned about how automated decisions are made, particularly for housing, financial services and employment, via what are termed high-risk AI systems due to the potential for consumer harm. A Colorado law requires developers of those systems to use “reasonable care” to avoid algorithmic discrimination, according to Bryan Cave, and includes transparency and auditing requirements for developers. Protections like those could help companies strengthen their contracts with AI vendors.
As financial institutions increasingly rely on AI, bank leaders need to understand what’s in the models they’re using. No one wants to run afoul of existing consumer protection laws, which are “tech agnostic,” Mahdavi explains. “If you make a discriminatory housing decision, whether that decision is made by a human or a machine, the law applies equally.”
Federal legislation doesn’t yet address these issues, she says. But as financial institutions deploy AI technology, some legal protections could help them work with their technology providers to balance compliance and trust with innovation.
• Emily McCormick, vice president of editorial & research for Bank Director
| 
