The One Thing AI Cannot Do for Your Students

More than 10 million annotations in a single academic year. Two consecutive semesters of over 5 million, a first in the history of Hypothesis, and it is happening at a moment when the question of whether students are actually thinking has never felt more urgent.

The timing is worth noting. The years following the pandemic brought a shift in higher ed — institutions that had rapidly adopted digital tools began to reassess their priorities, budgets tightened, and many platforms saw their momentum stall.Yet, despite that industry-wide pullback, the volume of social annotations on Hypothesis grew by 40% since 2022. This growth didn’t happen because of top-down mandates; it happened because faculty, facing unprecedented challenges in student engagement, choosing annotation as a core part of how they teach. It’s proof that when the noise is stripped away, educators double down on what actually drives deep reading and critical thinking.

And then AI changed the conversation entirely.

The challenge faculty are navigating right now is a specific one. AI can write the essay, summarize the reading, and produce a discussion post that sounds thoughtful in seconds.  The tools that used to require genuine effort no longer do. Students can move from assignment to submission without ever really engaging with the material.

This has shifted the question faculty are asking. It is no longer just “did my students do the reading?” It is “how do I know my students actually thought about it?”

Detection is not the answer. The tools evolve faster than any policy can keep up. What works is proof, the kind that comes from a student engaging with a text in a way that is specific, personal, and impossible to outsource. What did they notice? What confused them? What did they connect to something they already knew?

Annotation captures exactly that. It makes thinking visible at the moment it happens, in the margin, in response to a specific sentence, in a particular reader’s own words. That moment of a human mind encountering a text for the first time is something no model can replicate. Not because AI is not sophisticated, but because the value of that moment is precisely that it belongs to a specific person.

That is why annotation has held its ground through a period when a lot of other edtech has not. It does not just measure engagement, it creates it. And in a landscape where AI has made passive learning easier than ever, that is not a small thing. See how institutions are using it. Check our case study library.

Nearly 10 million annotations. The number is meaningful not because it is large, but because of what each one represents: a student who had to show up, read carefully, and leave a trace of their own thinking.

That is what higher education is trying to protect right now. And it is what social annotation is built to do.

Want to see Hypothesis in action?

Request a demo and we will show you how Hypothesis works in your LMS.

Get Started

FAQ

What is social annotation?

Social annotation is the practice of reading and annotating a text collaboratively, where students and instructors can highlight, comment, and respond to each other’s notes directly on the document. Unlike a discussion board, the conversation happens in context, right next to the words that sparked it.

Can students use AI to complete annotation assignments?

Not effectively. AI can summarize a text or generate generic observations, but it cannot replicate what a specific student noticed, felt confused by, or connected to their own experience in the moment of reading. Well-designed annotation prompts ask for exactly that kind of situated, personal thinking, which is why annotation holds up as an assessment even as AI tools get more sophisticated.

How is annotation different from a discussion board?

Discussion boards ask students to respond to a prompt after reading. Annotation asks students to respond to the text itself, in the moment, at the specific sentence or paragraph that sparked a thought. The result is a much closer record of how a student actually engaged with the material.

How do I get started with Hypothesis at my institution?

Hypothesis integrates directly with Canvas, Blackboard, Moodle, and other major LMS platforms. If your institution already has a subscription, you can get started right away. Request a demo and we will walk you through how it works in your specific setup.