Lawmakers wrestle with the ever-evolving role of AI in education, election integrity and its legal implications. Photo illustration by Canva.

By Whitney Downard
Indiana Capital Chronicle

INDIANA — Artificial Intelligence, or AI, promises to revolutionize how people work and nearly every aspect of life could be transformed — prompting lawmakers in an interim commerce committee to scrutinize the new technology and how to best regulate it.

“We’ve had the pants scared off of us a little bit, in some regards,” said Committee Chair Sen. Scott Baldwin, R-Noblesville, summarizing portions of the meeting.

In the months since they left the capitol, AI has exploded in popularity but regulation and guidelines haven’t accompanied it — forcing lawmakers to act faster than the traditional legislative process that can stretch years.

Following more than four hours of testimony, legislators landed on three areas of focus to include in a final report to the General Assembly concerning AI: a study of which currently applicable laws might need to be expanded to include AI, election interference and pivoting educational priorities to include artificial intelligence.

Background on artificial intelligence

Hoosier Justin Sheehan, of Astra Applications, likened AI to computers adopting humanlike thinking but emphasized repeatedly that he didn’t see it capable of replacing humans. Rather, it was the act of computers acquiring learning, reasoning and problem solving skills.

“AI feels like it’s this new thing that everyone’s excited about but actually AI has been a part of our lives for a very long time,” Sheehan said, noting that voice recognition tools like Apple’s Siri and spam filters are a type of artificial intelligence.

The emergence of ChatGPT and similar, generative models earlier this year pushed AI into the forefront due to its human-like responses, accessibility and effectiveness, Sheehan said.

“AI learns things like humans; it can create, generate content from scratch (and) it can execute complicated tasks. It can remember and understand context — which is important — has access to vast amounts of information (and) it’s always learning,” Sheehan said.

The technology can write content, plan weddings, tutor math or act as a personal assistant. But it has a myriad of problems, including biases — whether based on race or gender — as well as the spread of disinformation, cybersecurity and surveillance concerns.

“There’s certainly a ton of upsides but there are also risks,” Sheehan said. “It’s really right now (in) a gray area in terms of federal regulation, in terms of state regulation (and) in terms of business best practices.”

What laws does Indiana have on AI?

Though much of the discussion delved into ways AI evolution will change society, legislators focused on what actions they should take on the rapidly-changing technology. Following the passage of a data privacy law earlier this year, Indiana joins the growing number of states with comprehensive protections — a move praised by Ryan Harkins, a senior director of public policy at Microsoft.

“In our view, we need laws. We need laws to ensure both that the public can be (confident) that these sorts of technologies will be developed and used and deployed safely and responsibly and ethically,” said Harkins. “But also to make sure that everyone is playing by the same set of rules.”

Harkins outlined a five-point blueprint, first created by Microsoft President Brad Smith in May, for governing AI, including:

• Implement and build upon new government-led AI Safety Frameworks
• Require effective safety brakes for AI systems that control critical infrastructure
• Develop a broad legal and regulatory framework based on AI’s tech architecture
• Promote transparency and ensure academic and nonprofit access to AI
• Pursue new public-private partnerships to use AI as an effective tool to address the inevitable societal challenges that come with new tech

Examples could be making sure applicable laws include AI technology; outlawing the use of AI to create “deep fakes” and influence elections; regulating safety and security;  licensing for training and deployment of powerful AI models, as included in Harkins presentation.

Lawmakers worried about whether the creation or possession of child pornography generated by AI would be considered illegal under current law — one way that AI might be circumventing current protections.

Rep. Matt Pierce, D-Bloomington, said the current law should be broad enough to include AI content.

“We currently have a law that no matter how you generate it … if you create any kind of child pornography that doesn’t involve real children — you sketched it out, you use Adobe Photoshop to create something — however you get it, you can be punished for that so long as you can also get a jury to say it’s obscene,” Pierce said.

Transforming education

The committee will meet for the final time on Nov. 1 to discuss draft recommendations for the General Assembly.

Following Wednesday’s testimony, members discussed three broad areas to include in a final report: application of current laws to AI (specifically child pornography), addressing election issues and incorporating artificial intelligence into classroom learning.

Legislators heard at length from several educators about their use of AI and how it has transformed their teaching.

That includes incorporating humanities and the arts to spur critical thinking and ethical decision making, Rikus said. A “blanket ban,” she said, would be “unnecessary and probably ill advised.”

Just as technology cannot replace a teacher, students shouldn’t use AI to replicate assignments or cheat but find ways to incorporate the technology into their learning.