How Indiana government uses AI for some services
INDIANAPOLIS (WISH) — Indiana’s top information technology officials on Wednesday told lawmakers the state government is adopting artificial intelligence slowly to ensure technology doesn’t get out ahead of their ability to regulate it.
Wednesday marked the first time the state’s AI task force had convened since Republican Gov. Eric Holcomb signed legislation establishing it this spring. Lawmakers on the panel said their biggest concerns were the potential for AI to pose a data privacy risk.
Chief Information Officer Tracy Barnes said the state currently uses AI for three applications. One is a program that recommends employment and training resources to people who file for unemployment benefits. A second, still in the beta-testing stage, is a chatbot meant to help people connect with the services they need. The third is a security log program for internal cybersecurity uses. Barnes said state agencies are considering AI for document review and email summary uses, and the Department of Correction is looking into using an AI program to help monitor security camera feeds for unusual activity.
Committee co-chairs Rep. Matt Lehman, R-Bremen, and Sen. Liz Brown, R-Fort Wayne, questioned Barnes and Indiana Chief Privacy Officer Ted Cotterill extensively over data retention and cross-department access. Barnes and Cotterill said the state already has strict rules that govern what types of data can be retained and for how long.
“I think what’s different for us is that AI presents the ability of a computer to sort of be humanlike,” Cotterill said. “Where a traditional IT system, we’re used to interacting with that, the dynamic has changed. It feels much more like we’re interacting with a person.”
Cotterill said the main thing the state needs is a way to refine its protection processes and make them scalable, especially for outside vendors the state works with.
Speaking to reporters afterward, Lehman said Wednesday’s hearing provided a useful overview of the state’s current AI use. He said he expects future hearings will focus more on specific applications, particularly education and law enforcement. Rep. Matt Pierce, D-Bloomington, who is a university lecturer in addition to his legislative duties, said one concern he has is whether AI will help students learn better or merely give them a new way to cheat. He said he suspects the legislature won’t step in unless they find a glaring issue not addressed by existing guidelines.
“The question is, at what point does the government through laws or regulations put in place guidelines to ensure you maximize the positive impacts on society and you limit the negative impacts, and I don’t think we know what that mix is yet,” he said.
Barnes told News 8 afterward generative AI is simply a new tool to help humans complete an old task: sift through large amounts of data quickly and analyze it. He said most of the rules and regulations AI needs for data protection already are in place.
“What we’re putting in place is pure protection of our data and the privacy of our data and the technology that we’ve been employing for many, many years,” he said. “There’s not really a need for new guidelines as much as, how do we make sure that AI solutions and AI vendors are being held to that standard?”
The AI task force will have to provide any legislative recommendations no later than Oct. 31.