Texas Rep: Collaboration, Privacy Basis Needed for AI Regulation
Regulation of AI tools and systems is all based on the same data governance principles used in privacy law, and it’s important for this regulation to be tackled collaboratively on a state level, said a Texas legislator and privacy and emerging tech experts during a regulatory panel at the Association of National Advertisers (ANA) AI and Technology for Marketers Conference Friday.
Sign up for a free preview to unlock the rest of this article
“Whenever anyone talks to me about artificial intelligence, I make sure to point out that AI is the next step of data,” said Texas Rep. Giovanni Capriglione (R). “This is all about information that's being collected from consumers,” which means that “what we want to make sure -- everybody needs to make sure -- is that we are…responsible and ethically handle and manage that data.”
There is a balance to be struck “because this data isn't just your social security number, just your birthday; some of it is very personal information,” he said. But “at the same time, consumers expect ads that make sense to them. They expect relevancy," said Capriglione: When they go on social media or other websites, "they're expecting to get stuff that's tailored and meaningful to them and their family.”
ANA Executive Vice President Chris Oswald agreed. “You really need to know where your data is coming from, what is the permission [of data use] for, and making sure that the AI that you're using training on that data doesn't run afoul of existing law.”
Capriglione said it was state legislators’ goal when passing the Texas Consumer Privacy Act to “make sure that consumers have rights to know what information is being collected, to be able to give that consent, to be able to correct that information, to sometimes ask for it to be deleted, and also not to be discriminated against if they don't want that data being collected.”
Chris Duffey, head of GenAI Studios at Adobe, said this approach “highlights the complexity and the need for finding that balance between utility [and] privacy.”
With his proposed Texas Responsible AI Governance Act (HB-149), Capriglione said they started by “getting working groups together, saying, ‘Okay, we're going to do something on artificial intelligence, let's say in a year, if you're interested, come on by.’”
“As part of that process, we ended up ... getting just over 300 documents of different kinds, giving us ideas, suggestions everywhere from" changing a conjunction to "just delete the whole entire bill,” he said. “We worked on it, and I think where we are now is the culmination of the opinions of a lot of different people, but also with the goal in mind of protecting our constituents here in Texas.”
Collaboration is also key across the country, said Capriglione, who was part of a state lawmaker AI working group. In addition to getting input from stakeholders, state lawmakers “have an opportunity to work with other states, see what's working, what's not” when drafting this legislation, he said.
“Nobody wants to see a patchwork of anything,” said the Texas lawmaker. “We would all like there to be consistency and rules of the road in every single state, with obvious differences. As a state legislator, I do believe in state rights, I do believe that the federal government should be limited in its scope. But that being said … I think we're going to get it right.”
“What we're trying to do on the state level, even though it looks like chaos, [is] that we talk to one another,” he said. “We're trying to make sure that our definitions are the same across states, we're trying to make sure that at least the intents are the same across state[s].”