Texas Lawmakers Think about Rules to Handle Dangers of Synthetic Intelligence

A gaggle of state lawmakers started wading by way of the advanced world of synthetic intelligence Tuesday, offering an early glimpse at how Texas could search to control the booming expertise.

Throughout an almost four-hour listening to, the Texas Senate Enterprise and Commerce Committee heard a variety of issues concerning the potential dangers of AI, together with the unfold of misinformation, biased decision-making and violations of client privateness. By the tip of the listening to, at the very least among the 11 committee members appeared satisfied that the state ought to enact legal guidelines that regulate how and when non-public firms use synthetic intelligence.

“If you happen to actually give it some thought, it’s a dystopian world we may stay in,” Sen. Lois Kolkhorst, R-Brenham, mentioned throughout the listening to. “I believe our problem is, how can we get on the market and put in these safeguards?”

Synthetic intelligence is a broad time period that spans a spread of applied sciences, together with chatbots that use language processing to assist reply customers questions, generative AI that creates distinctive content material and instruments that automate choices, like how a lot to cost somebody for house insurance coverage or whether or not a job applicant ought to get an interview. Synthetic intelligence can be used to provide digital replicas of artists’ work.

Already, greater than 100 of the 145 state companies are utilizing AI in some type, Amanda Crawford, chief data officer for the Texas Division of Info Sources, advised lawmakers Tuesday. Crawford is a member of a brand new AI Council created this yr by Gov. Greg Abbott, Lt. Gov. Dan Patrick and Home Speaker Dade Phelan. The council is tasked with learning how state companies use AI and assessing whether or not the state wants a code of ethics for AI. The council is anticipated to publish its report by the tip of the yr.

Leaders of a number of state companies testified that synthetic intelligence has helped them save important money and time. Edward Serna, government director of the Texas Workforce Fee, for instance, mentioned a chatbot the company created in 2020 has helped reply 23 million questions. Tina McLeod, data officer within the Legal professional Normal’s Workplace, mentioned their employees have saved at the very least an hour per week with an AI software that helps sift by way of prolonged little one assist instances.

However in different instances, stakeholders testified, AI expertise may very well be utilized in ways in which damage Texans.

Josh Abbott, a rustic singer, mentioned he worries AI may very well be used to copy his voice and generate new songs that get distributed on Spotify.

“AI fakes don’t care if you happen to’re well-known,” Abbott mentioned. “AI frauds and deep fakes have an effect on everybody.”

Grace Gedye, a coverage analyst for Shopper Studies, recounted how non-public firms have already used biased AI fashions to make essential housing and hiring choices that damage customers. She mentioned lawmakers may mandate firms who depend on AI for decision-making to audit that expertise and speak in confidence to customers how they’re being evaluated.

Gedye pointed to New York Metropolis, which enacted a legislation requiring employers who use automated employment instruments to audit these instruments. Few employers truly carried out the audit, Gedye mentioned.

In creating laws, the state might want to tread fastidiously and ensure they don’t write legal guidelines that inadvertently prohibit constructive makes use of of synthetic intelligence whereas they attempt to clamp down on harms, mentioned Renzo Soto, an government director of TechNet, which represents expertise CEOs.

“You nearly have to take a look at it trade by trade,” Soto mentioned.

Texas already handed a legislation in 2019 that makes it against the law to manufacture a misleading video with the intent to affect an election. Final yr, lawmakers handed one other legislation prohibiting using deep faux movies for pornography.

After they think about future laws to rein in synthetic intelligence, lawmakers will want to ensure they don’t violate First Modification free speech protections, mentioned Ben Sheffner, an lawyer for the Movement Image Affiliation.

All through the listening to, lawmakers repeatedly requested whether or not different states or nations may very well be checked out for templates on how you can craft AI coverage. Thus far, a patchwork of state and federal laws have tried to restrict AIs use, with restricted success.

California lawmakers have launched a invoice that requires AI builders and deployers to cut back the dangers of “catastrophic hurt” from their expertise. Tech firms are preventing to kill that legislation. Colorado has additionally handed a legislation regulating using AI in sure “high-risk” situations, comparable to these pertaining to training, employment or well being care. Colorado’s governor has already mentioned the legislation must be amended earlier than it goes into impact in 2026.

This text initially appeared in The Texas Tribune at https://www.texastribune.org/2024/08/27/artificial-intelligence-regulation-texas/.

The Texas Tribune is a member-supported, nonpartisan newsroom informing and fascinating Texans on state politics and coverage. Study extra at texastribune.org.

Matters
Texas
Legislation
InsurTech
Data Driven
Artificial Intelligence