Listen to the article
This audio is auto-generated. Please let us know if you have feedback.
As artificial intelligence tools continue to rapidly evolve, some K-12 leaders say it’s time to switch up districts’ approach to their adoption and implementation.
Unlike how districts commonly procured ed tech in the past, decisions on AI tools need to be reevaluated over time, said Julia Rafal-Baer, CEO of ILO Group, an education strategy and policy firm, and education consultant.
“It’s an ongoing leadership practice, which means leaders must now know if they have the governance structures and the organizational capacity to learn quickly, make decisions responsibly and adjust as the technology evolves,” Rafal-Baer said during a Monday webinar about ILO Group’s March report on AI governance in schools.
There are multiple ways to go about this work, she said. For instance, school leaders can create AI leadership roles at the district level, integrate AI responsibilities into existing district teams or distribute ownership across the whole school system.
To understand how education leaders are already doing this work, the ILO webinar highlighted two school districts and one state level case study as examples of these practical frameworks for adaptive AI governance in action.
Charlotte-Mecklenburg Schools’ AI governance team
Before Charlotte-Mecklenburg Schools in North Carolina began to roll out its policies and AI governance team, Superintendent Crystal Hill shared during the webinar, her district surveyed the school community about using AI tools and received about 10,000 responses.
Then after the district worked with its legal team and school board members to develop a districtwide AI policy. And finally, Charlotte-Mecklenburg Schools formed an AI governance team that includes staff from instruction, assessment, technology and principal leadership departments, she said.
“That really helped us take a look at all of the things that our community had said and drive that into a very clear strategy,” Hill said. From there, the district launched a pilot program to test AI tools in 30 of its 185 schools — instead of letting everyone in the district experiment with AI on their own.
Before the district enabled AI tools in a school, Charlotte-Mecklenburg trained its 14,000 staff members, Hill said. All 140,000 students have completed digital citizenship and literacy training as well, she said.
Hill said her district is pushing to focus on using AI in these pilot programs as a way to solve real problems and learn from each use case. “It’s making sure that we’re really scaling what is and isn’t working.”
Indianapolis Public Schools’ districtwide AI approach
For Aleesia Johnson, superintendent of Indianapolis Public Schools, it’s important that AI is not a siloed initiative or program. Bringing the district’s technology leader to the decisionmaking table with Johnson’s leadership team has helped her keep up with the pace of the technology’s rapid development, she told the ILO webinar.
Additionally, Johnson said, a district needs to use a continuous improvement cycle for implementing AI in schools after it finalizes its policy on AI use. Districts need to gather feedback and adjust how they approach AI with that in mind, she said.
Delaware’s AI assurance lab
Implementing AI tools requires a repeating cycle of testing for evidence of a tool’s success and correcting course as needed, according to Cynthia Marten, Delaware’s Education Secretary. However, not every district has the capacity to continuously test a lot of AI tools, she said during the ILO webinar.
That’s where the Delaware Department of Education is stepping in through its AI assurance lab, which tests AI tools for K-12 use within state set guardrails, Marten said. The lab supports AI procurement guidance for districts statewide, she said, by gathering data from teachers as they try different tools. The lab can also highlight for districts any best practices or early concerns about AI tools as they emerge.
Through the lab, the state is helping districts quickly test and evaluate AI tools in a safe way that prioritizes teacher voice and efficacy for student outcomes, she said.
“We have benchmarks for what AI can do,” Marten said. “We don’t yet have enough evidence for what it does for learning, and that’s the gap states have to own.”
For other districts and states looking to responsibly lead on AI, Marten advised that they not slow down on implementation out of fear. “Have clarity, have consistency [and] a relentless focus on outcomes for students. Stay off the hype.”
Instead of trying to find the perfect AI tool, districts need to first understand what problems they want the technology to solve, Marten said. They should start by establishing clear guardrails as well as a cross-department team focused on AI, she added.
“We can move quickly. We can innovate. We can be fast in education, despite what people think,” Marten said. “This is about building the muscle to decide, learn, [and] improve in real time.”
