South Dakota lawmakers heard warnings about the risks of artificial intelligence, ranging from its impact on youth to data poisoning, as an interim legislative committee seeking to study the topic met for the first time.
Much of the world is already integrated with artificial intelligence. Examples can be found in text message auto-correct, bank alerts for possible fraudulent charges, and developing probabilities to determine future actions.
Lawmakers met with multiple state and national tech professionals this week to get a grasp on potential AI threats.
Joe Graves is the state secretary for the Department of Education. He said state-level regulation won’t be enough.
“It seems unlikely that our local or even our state responses to AI will have much of an impact on just how AI plays out in our society and our world. This looks to me to be a tsunami, and a fast-moving one. And tidal waves are notoriously hard to predict or prepare for,” said Graves.
Graves said the Associated School Board of South Dakota has drafted a model policy for AI which all school districts in the state have adopted.
Jose-Marie Griffiths is the president of Dakota State University. She said one of the largest areas of concern with AI is what her team has been studying: data poisoning.
“Let’s say a traffic signal is a stop light or a go. If you wanted to, you could break into that signal that’s going from a post to an electric vehicle, you could change it. So, something could go through what’s really meant to be a red light, and do some damage,” said Griffiths.
Griffiths said hacking electric vehicles is just one area data poising could affect the world.
“You could do a lot more damage by going to the GPS system, and just tweaking it a couple of degrees, literal degrees and you would have chaos everywhere because everybody would be just off enough. And that can be done,” said Griffiths.
Tech industry representatives agreed regulation should be made by each state. They also warn lawmakers, however, about going too far and stifling innovation.
Heather Morton is an analyst for the nonpartisan National Conference of State Legislatures. She said it is hard to keep up with the over 300 AI-related bills being filed across the nation.
Morton said legislation aimed at social media and children cover a wide variety of issues, including age-appropriate design code, parental consent and supervision, harmful material age verification, media literacy and protecting content creators.
Each avenue of social media presents its own risks and challenges.
Morton said examples can be found from 30 states that enacted social media regulations this year.
“Some of the examples of the bills that have been enacted this year include Colorado, which has required that the state department of education create and maintain a resource bank of existing evidence-based, research-based, scholarly articles and promising programing materials and curricula that pertains to the mental and physical health impacts of social media, used by youth as well as internet safety and cybersecurity,” said Morton.
Morton said one of the newer areas of regulation is protections for young content creators.
“States are also looking at the labor and compensation protections for children that are content creators. They create vlogs, video blogs, or may be seen as kid influencers. For example, last year, Illinois enacted the first kid influencer of 'vlogging law,'" said Morton. "Then Minnesota followed this year to make sure that parents are actually providing to compensation for minors who are appearing in those internet content creations.”
Rep. Mike Weisgram is the co-chair of the summer study. He asked Morton what type of challenges are being presented by the tech industry on these limitations and issues.
“Do any of the social media companies, are they helpful at all when some legislation being desired to be drafted? I mean, they’re parents and grandparents too right?" asked Weisgram.
Morton answered by saying social media companies are lobbying directly on these bills but that she doesn’t know in which ways they support or disagree.
The next summer study committee meeting on artificial intelligence and regulation of internet access by minors is Aug. 14.
