What Ally Can (and Can’t) Tell You About Your Course

Online learning is one of the best tools we have for expanding who gets to participate in higher education. Whether a student is managing chronic illness, caregiving responsibilities, geographic distance, or a disability that makes commuting to campus difficult, the ability to learn asynchronously — on one’s own schedule, from their own space — isn’t just a convenience for students… it’s the thing that makes participation possible at all. That population disproportionately includes students with disabilities as defined under the ADA and Section 504, which is exactly why the accessibility of what’s inside our online courses deserves as much attention as the fact that the course exists online in the first place.

The access argument isn’t just philosophical anymore, either. Now it has a compliance clock attached to it: the Department of Justice’s Title II rule requires public colleges and universities to bring web content, course materials, and third-party tools up to WCAG 2.1 AA. Though the DOJ pushed the original April 2026 deadline back a year (to April 2027 for larger institutions, 2028 for smaller ones), the underlying nondiscrimination obligation hasn’t gone anywhere in the meantime. If your institution is public, this isn’t a someday problem, it’s a “start now, because testing and remediation take longer than people expect” problem.

That’s where a tool like Ally comes in. If you’ve worked in or around an LMS in the last several years, you’ve probably seen it. If you haven’t, the short version is that Anthology’s (aka Blackboard) accessibility-checker integration that scans content, assigns each file a score, and generates a plain-language report on what’s wrong and how to fix it. It can also provide whole-course or whole-department/campus reports, depending on how an institution’s LMS admin has it configured. If you want the vendor’s own overview of what it does, Anthology has documentation here.

I’ve spent a lot of time reviewing Ally reports this year, both as a QA/remediation consultant and as someone who just cares a lot about the user/student experience. Here’s what I’ve learned about where it earns its keep, where it quietly runs out of road, and what that means depending on which chair you’re sitting in.

Where Ally genuinely helps

Ally is good at the job it was built to do: catching structural and technical accessibility issues in content that lives directly inside the LMS. Missing alt text, poor heading hierarchy, low color contrast, PDFs and Word docs with no underlying tag structure — Ally flags all of it clearly, and in some integrations, you can fix the simple stuff (like adding alt text to an image) right there in the report interface without ever leaving the review screen.

It’s also, maybe more importantly, a really effective mirror held up to a pattern I run into frequently in my work: content-dumping. When a course goes from face-to-face to online, the instinct is often to upload the same files instructors have always used — lecture slides, handouts, readings, etc. — without adjusting them for a format where there’s no live instructor filling in the gaps. Ally’s scoring often forces a second look at that content, and that second look is often begs the question, “does this file actually work as a standalone learning object?” rather than just “is this file present?”

A good example from my own work: a set of lecture slide PDFs in an asynchronous course, originally built for in-person delivery, had zero accompanying commentary. Students were left staring at a contextless outline — bullet points that clearly left a lot of critical insights unexplained. Turning those PDFs into accessible, standalone HTML pages wasn’t just an accessibility fix; it gave the instructor a natural pathway to add the narrative and context that had previously only existed in their head, or in a room the online students never got to be in.

Two more concrete examples from that same vein of work: a complex flowchart buried in a PDF became an interactive, screen-reader-navigable SVG that students could actually explore rather than squint at. And a full history of art course’s slide decks became interactive, accessible slideshows — hosted entirely within the LMS — complete with descriptive alt text, so Ally (and any screen reader) could read the visual content aloud in a way that actually conveyed the art, not just “image.”

If you want to see the before/after on a couple of these, I wrote up a full case study here, including a live demo of a PDF converted into an accessible HTML page.

Where it runs out of road

Ally’s real limitation isn’t a flaw exactly — it’s a boundary. It can only meaningfully evaluate what lives inside the LMS: files uploaded to course content areas, text written directly into pages. When it comes to content that is linked out to an external site, or embedded via iframe, Ally’s visibility gets much thinner. In my work, this has affected critical content like journal articles, publisher platforms, and external tools, which instructors have come to rely on. All of that lands in this literal blind spot.

Journal articles are the one I hit most often when remediating courses. They’re the core currency of knowledge in higher ed, and the formats they’re traditionally published in can be a mess for screen readers, especially older, seminal literature — it’s dense, unstructured PDFs with no semantic tagging. ID professionals are generally inclined to solve this by linking directly to the publisher’s platform instead of uploading the PDF, and hoping the browser can handle it, or that an alternative format is offered. Sometimes that works, but it introduces problems too, like link rot and persistence over time, and no guarantee the external site itself is accessible.

To be fair, publishers are aware of this and slowly moving on it. Taylor & Francis has converted a large chunk of its catalog to EPUB3 (an accessible format) and is beginning to roll out AI-generated alt text for journal images (starting this year in 2026). But even with real, publicly-committed progress, that still only reaches a fraction of a catalog with decades of legacy content behind it — and AI-generated alt text is a floor, not a ceiling, especially for anything more visually complex than a photo. The ownership and reproduction questions around remediating copyrighted material for accessibility purposes are also unclear, and in a lot of institutions, this whole tangle gets punted to library staff to sort out article by article. Suffice to say this is something our field is really grappling with, even with deadlines looming.

The scoring nuance that matters

Here’s what I want everyone who looks at these Ally reports to understand: a high Ally score is not the same thing as a genuinely accessible course. It’s a proxy, and a useful one, but a proxy for what’s stored in the LMS specifically. A course can score 99 while linking out to external content that’s completely inaccessible, and Ally will never see it, because it was never designed to evaluate the open web.

This score says something different depending on who’s reading it:

• If you’re an instructional designer: you already probably know this, but it’s worth saying plainly to the people who don’t — a high score is a floor, not a ceiling. Keep asking what’s linked out and iframed, not just what’s uploaded, and consider whether WCAG 2.1 AA compliance is enough, or if a document needs a different treatment to be truly useable for students.
• If you’re teaching without ID support: you’re not expected to be an accessibility expert on top of running your course day-to-day, and that’s exactly why a tool like Ally exists — but treat the score as a starting checklist, not a finish line, especially for anything you’re linking to rather than hosting directly.
• If you’re in a leadership or administrative role: a department-wide Ally score is a useful signal, but it needs to be read with context. It tells you a lot about your internal content and very little about the quality of the broader course experience. If your institution doesn’t have dedicated ID staff to decipher that nuance, that’s a gap worth addressing — either by investing in ID capacity or bringing in outside support for a defined project.

Where this leaves us

Ally does real, valuable work, (this post doesn’t even touch on what the tool does for students, which is genuinely cool and very useful) and it’s one of the better nudges the field has for getting faculty and institutions to actually look at their content instead of assuming it’s fine. But it’s a starting point for a conversation, not a substitute for someone who understands both the technical and pedagogical sides of what “accessible” actually means for a given piece of content. That’s the gap I spend most of my consulting hours in — the space between “Ally flagged it” and “this is actually usable, useful, and good.”

If you’re navigating an Ally rollout, a compliance deadline, or a backlog of legacy PDFs your faculty have been quietly dreading, I’d love to talk. You can see more of my approach in my accessibility QA & remediation case study, or reach out directly.

Further reading: for a research-based look at how faculty and staff perceive using Ally in practice, see this Open Praxis article.