Education leaders at the Bridges 2026 conference in New York City on July 22 disclosed that K-12 districts and colleges are running AI agents in production for enrollment processing, student tutoring, and administrative workflows, according to GovTech.
The biggest divide among panelists was not whether agents have value, but how much system access and autonomy to grant them.
Two Approaches to Agent Autonomy
Milwaukee Area Technical College has intentionally restricted its agents to faculty-controlled tutoring through Gemini Notebook. Sadique Isahaku, the college’s vice provost and chief academic officer, told the conference the institution is cautious about granting agents broader access.
“We are very, very careful about agents because they have potential to get into your systems, and maybe, we don’t know yet … collect data that probably shouldn’t go out to the public,” Isahaku said.
Johnston Community College in North Carolina took the opposite approach. Carrie Pitts-Densmore, the college’s vice president of enrollment and marketing, said agents are integrated into administrative workflows that access live student records. The college could not hire staff fast enough to match enrollment growth, so agents now move students through the enrollment cycle from inquiry to completion.
“We couldn’t hire as quickly as we were growing, and so in order to scale that, we had to put these agents in motion,” Pitts-Densmore said. She noted that during the conference itself, she had 12 AI agents performing tasks on her behalf.
The Verification Layer
Isahaku described a deployment pattern for high-stakes work: agents handle the repetitive processing of monthly enrollment reports that determine institutional funding, while humans verify the final output.
“When we’re talking $150 million a year, $60 million a year, you don’t want to screw up those numbers … you do want somebody just at the very end to verify those before we file them with the state,” Isahaku said.
Kris Hagel, CIO of Washington’s Peninsula School District, said his district piloted agents last year and plans to make them available to every adult staffer this fall. The current work focuses on building the skills assigned to agents, according to GovTech.
Adoption Framing
Panelists agreed that deployment is no longer the primary challenge. Preparing institutions, staff, and students to use agents effectively is. Pitts-Densmore emphasized that leaders need internal champions and a culture that permits failure during experimentation.
Joe Siedlecki, chief impact officer at Amira Learning, drew a line between agents and chatbots during the panel: a real AI agent “makes decisions pinned back to a specific goal” and adjusts based on observed outcomes, rather than responding to individual prompts.
The education sector’s adoption pattern mirrors a broader dynamic visible across enterprise deployments. The technology works. The governance question, specifically how much access and autonomy to grant, is where institutions diverge.