Here’s the text of an internal email that I sent today at my college, and that touches on issues of general interest outside the walls of academia, which is why I’m sharing it here with some details changed to remove people’s names. It deals with the sudden, awesome challenge of agentic AI inside higher (and other) education institutions, as tied to our use of online learning management systems:

I'm sending this to our core academic leadership, assessment, and LMS team. Attached is a striking piece in The Chronicle of Higher Education showing Blackboard's chief product officer saying the company has found that it's essentially impossible to create tools that will detect the use of AI agents in an ethical way, and that they're therefore switching their emphasis to building tools into Blackboard that will help faculty and institutions bring more authentic and reliable assessment practices to their classes. In other words, they're kind of throwing in the towel on direct AI detection technology.

This subject is not an idle or purely academic one for us. For those who don't know, one of our faculty remembers recently found that one of her students was directing other students to SolveMyBook, a paid academic "assistance" service that's actually an agentic AI product designed to interface with McGraw Hill textbooks and course materials. It acts on a student's behalf inside a school's LMS by completing work for them, including everything from typing text to clicking buttons and more. It's also specifically designed and tuned to mimic the behavior, rhythms, patterns, and pauses of a human, so that the difference is essentially undetectable on the receiving end. 

Also in case you hadn't heard, last month saw the launch of Einstein, an all-purpose autonomous AI agent created for completing work in Canvas. Inside Higher Ed ran an article titled "Agentic AI Can Complete Whole Courses for Students. Now What?" Here are the opening paragraphs:

Earlier this week, Advait Paliwal—a 22-year-old tech entrepreneur who dropped out of the computer science master’s program at Brown University in 2024—launched Einstein, an agentic AI tool specifically designed to connect with the popular learning management system Canvas.

“Einstein is an AI with a computer,” the product website explained when it first went live a few days ago. “He logs into Canvas every day, watches lectures, reads essays, writes papers, participates in discussions, and submits your homework—automatically.”

Einstein can help with any subject, including math, physics, computer science, history, literature and economics, and even keeps working when students are asleep.

While such promise may be alluring to some overwhelmed or unmotivated students, Einstein’s release this week is intensifying discussion among faculty about policy, pedagogy and the purpose of higher education in the age of agentic AI. 

That last sentence about the way this is all "intensifying discussion among faculty about … the purpose of higher education in the age of agentic AI" is massive in its import. It dovetails, I think, with some of what the Blackboard exec said to The Chronicle about the fact that this is all forcing us to reconsider assessment and find ways to make it more authentic and accurate for measuring student learning. The whole thing reminds me of the growing argument many people are making about the way the threat of AI to displace people professionally and creatively in a multitude of fields is forcing a sudden new reckoning with and clarification of what it means to be human, and what the qualities and characteristics are that define not just work but us, in ourselves, as human, as persons.

It's a fraught moment. The philosophical implications lead right into the practical threat/challenge we're facing at this very moment. I've attached the Chronicle article with my highlights added. I encourage you to read it. It's worth bringing all this to our ongoing conversations.

(To readers of this Substack Note: If you want to read that Chronicle article, look up “Blackboard Executives Say Catching AI Cheating Is a Lost Cause. This One Isn’t Worried.” It’s behind a paywall, though.)