California’s top consumer regulator is signaling a potentially far-reaching approach to policing artificial intelligence: If state law requires a human to hold a license to perform a service, deploying AI to perform the same function may not provide a way around that requirement.
Rohit Chopra, secretary of California’s new Business and Consumer Services Agency, said the agency will investigate businesses’ use of AI, chatbots and other technologies, including whether some applications violate prohibitions against the unlicensed practice of medicine or other professions.
While healthcare provides the most obvious target, the approach could reach financial services, real estate and other regulated professions overseen by BCSA, potentially turning longstanding licensing laws into another tool for regulating increasingly autonomous AI systems.
“The Business & Consumer Services Agency will investigate how businesses are using new technologies to ensure that they aren’t harming consumers or undermining longstanding legal protections,” Chopra wrote in an Aug. 31 blog post. “For example, we will be examining whether certain uses of these new technologies may violate prohibitions on unlicensed practice of medicine or another profession.”
The approach effectively shifts the regulatory inquiry away from AI itself and toward what the technology actually does. Rather than waiting for lawmakers to establish AI-specific restrictions, regulators could ask whether an activity performed by software is already reserved by law for licensed professionals.
That distinction could become increasingly important as businesses move from AI copilots that assist employees toward agentic systems capable of making decisions and acting with limited human supervision.
Healthcare provides a ready-made testing ground. California already has laws restricting AI systems from representing themselves as licensed healthcare professionals when they are not. State law effective this year extended existing restrictions on falsely claiming professional credentials to AI technology providers, with healthcare licensing boards able to enforce violations, according to a tally of state measures kept by Orrick’s AI Law Center.
But Chopra’s announcement points toward a potentially broader question: Can an AI actually perform functions that the state has reserved for licensed professionals, regardless of how the technology represents itself?
The implications could be particularly significant for financial services.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
BCSA houses the Department of Financial Protection and Innovation, which regulates banks, credit unions, financial advisers and other financial companies. The agency also encompasses the Department of Real Estate and the Department of Consumer Affairs, whose boards and bureaus oversee numerous licensed businesses and professionals. Its jurisdiction reaches sectors including financial services, healthcare, real estate, retail, agriculture and professional services.
Orrick said Chopra’s announcement signals scrutiny of AI and chatbot deployments across sectors, including financial services, healthcare and real estate, as BCSA examines whether businesses are using AI to engage in activities requiring licenses.
For financial firms and FinTechs, that could make the regulatory perimeter surrounding an AI system as important as familiar concerns involving privacy, discrimination, explainability and data security. An artificial intelligence application that merely supplies information could present different licensing issues from one that independently performs regulated financial functions or exercises judgment traditionally assigned to licensed personnel.
The same analysis could apply to real estate and other professions.
Companies deploying advanced AI may need to map individual system capabilities against professional licensing statutes, determine which decisions require human involvement and examine whether third-party AI vendors are effectively performing regulated activities on their behalf.
The initiative also fits Chopra’s broader strategy for the newly created BCSA. In July, he said the agency would concentrate audits and inspections on entities posing the greatest consumer risks and stressed that existing state laws can carry consequences including loss of a license to operate in California. Some BCSA departments can also enforce federal law.
That suggests California’s AI enforcement regime may emerge partly through ordinary supervision and licensing rather than through a single comprehensive AI law.
It also offers a model other states could follow. Professional licensing statutes already exist across the country, potentially giving regulators an enforcement mechanism that does not depend on legislatures adopting new AI laws.
As AI evolves from providing answers to performing professional tasks, California is beginning to confront a basic regulatory question: Software may not need a diploma or professional credential, but does that mean it can legally do a licensed professional’s job?
Add as Preferred Source
