Kill Canvas. Now.

An outage at the popular LMS calls attention to higher ed's over reliance on technology

May 12, 2026

Today, we would like to welcome new subscribers Gabrielle Schleijpen, Jane Clark, Shireen Singh, Mike Barnhill, and premium subscriber Joyce Sherry.

Image credit: Antonio Guillem/Shutterstock

Last Thursday, the thinkable happened: the hacking group ShinyHunters took control of Canvas, a widely used learning management system (LMS) and held it for ransom.

Although everything was back up and running by Saturday, valuable time had been lost—for students, for faculty, and for the deans who determine who will get degrees this week. About 40% of educational institutions in the United States use Canvas; another 59% use another LMS like Google Classroom or Black Board. Combined, these platforms have over 73 million users, roughly the same as Pinterest, at its peak, and the number of users who log onto Reddit every day. Most of them are under the age of 21.

Yes, I wrote “thinkable” in my lede, and it isn’t one of my normal typos. So dependent, for so many reasons, has higher ed become dependent on LMS’s that blowing up exam period across the nation was an inevitability—irresistible, in fact, even if money had not been involved. According to CNN,

This was the second school data breach claimed by ShinyHunters this month. In Thursday’s ransom note, the group claimed it had hacked Instructure “again” and faulted the company’s response to the previous attack: “Instead of contacting us to resolve it they ignored us and did some ‘security patches.’”
ShinyHunters claimed in a ransom note shared on May 3 by Ransomware.live, which tracks ransomware attacks and groups, that it had breached 275 million individuals’ data and had access to “several billions of private messages,” giving a May 6 deadline for Instructure to reach out.
In a note Thursday, the hacking group gave a May 12 deadline for impacted schools “to negotiate a settlement.”
During the Canvas interruption, Instructure said on Thursday it put the platform in “maintenance mode” as it investigated the issue. Later that night, it announced Canvas was available again “for most users.”

Before we continue, let’s put a pin in two things. One is that we now know that Canvas successfully negotiated with the hackers to get the data back, which means they have paid for it. That’s right, sports fans: your tuition and tax dollars are now sitting in a crypto wallet connected to a crime kingpin somewhere.

What we don’t know is how much was paid. That isn’t surprising, since if your company is known for being willing to part with a sizable sum, , why shouldn’t another group of enterprising hackers give it a shot too?

Thing two is that Canvas has known since the end of April that they had a problem, and they warned no one. The hack was only a surprise to their millions of customers, who do not just attend and work at colleges and universities, but thousands of high schools. Nearly 9,000 institutions were affected, with billions of records extracted: wouldn’t a warning to faculty and students to back up Course materials, exams, and grades have been ethical?

College professors, high school teachers, and all of their students have been hostages to Canvas, other learning management systems (LMS’s), and technology more broadly, for over a decade. Yet this month’s crisis, which was cleverly timed to coincide with the few weeks a year when grades absolutely must be submitted so that graduates can move on to jobs, has a much longer history. Our journey to this moment began decades ago, with the first broad-based technology “solution” to reach education in the 1990s: .edu email accounts that then became “free” in 2006 when Google launched its suite of education apps.

Arguably, the technology tail started wagging the educational dog even earlier, when Apple targeted secondary schools as early adopters. “In 1978, just two years after it was founded,” tech journalist Audrey Watters wrote on her blog, Hack Higher Education, in 2015:

Apple won a contract with the Minnesota Education Computing Consortium to supply 500 computers for schools in the state. MECC had developed a sizable catalog of educational software (including the iconic Oregon Trail) which it made freely available to Minnesota schools. Soon the MECC floppy disks and Apple II’s became popular elsewhere across the country. As Steve Jobs said in a 1995 oral history interview with The Computerworld Smithsonian Awards Program, “One of the things that built Apple II’s was schools buying Apple II’s.”

Let’s be clear: educational technology “solutions” were never about delivering a better education. They were about selling hardware and software. If computerization of all things delivered more and better education, then why do we presently believe—as we reach peak tech with AI—that students read less, cheat more, write worse, and miss more deadlines than they did even ten years ago?

Crucially, LMS’s were also sold as a way that schools could reduce reliance on human employees, reach more students without building and maintaining more campus, cutting costs for educational materials (books, and what we used to call “course packs,” xeroxed chapters and articles that violated copyright six ways to Sunday), and monitoring learning outcomes.

I would say that the results of implementing technology “solutions” have been mixed at best, but they have also been destructive in the ways all new products—whether we are talking about breakfast cereal or DoorDash—which is that they solve invented problems by destroying something else. In the case of edutech, technology solutions have perhaps had their greatest impact in destroying community and attention spans, two things that many of us valued about our work when we went into it, and that students used to rely on to develop as human beings.

When email was introduced at Wesleyan University, where I began my career in 1991, we had a social networking system called “the departmental mailbox.” One’s day began and ended by checking the mailbox. I touched down at the mailbox after lunch; if teaching required travel to another building, a pitstop at the mailbox naturally followed. In a small cubby with my name on it, I received missives wrapped in envelopes. They could be crap. More interestingly, in terms of the networking thesis, there might be a memo that I was supposed to check my name off of (to indicate I had read it) and then circulate to the next person on the list (there were special yellow envelopes for this.) There might be a notice about a fellowship, a prize, a newsletter from my professional association, a student paper (late or not late), a box of my favorite pens ordered by the department administrator, or the agenda for a department meeting. In the fall, there was sometimes an anonymous zucchini or two.

In the world before technology provided a “solution” to every educational need, there were also people, invariably women, who actually supported your work. One job they did was sort the mail (delivered in a bundle by yet another worker whose job no longer exists.) These women were founts of deep knowledge about the department, and far more valuable to a new faculty member than any orientation day. They taught us how to get our job done. They let students know that the new prof’s seminar wasn’t full and they should try it. They explained why some machine or another was not working. They reassured weeping students that things would be ok and unlocked the office door at 5:05 so that a paper wouldn’t be counted as late. One of those secretaries stayed until 8:00 pm on a June evening to make sure my tenure file was perfect before I handed it over to the department chair.

There were also random encounters with other faculty at the mailboxes. Casual conversation there knitted the department together, even when the issues we faced as a collective divided us. It was an opportunity to be polite and cordial to people who, perhaps, you really didn’t like or really didn’t agree with, but needed to forge a working relationship with.

Can you convert to paid?

Email smashed that system, and it is a microcosm of the changes sold to us as improvements to higher education that, in reality, saved money for the institution by transferring work to faculty desks. A senior colleague of mine predicted this the day that using our email accounts became mandatory. “It’s speed-up,” she said in 1993, a process by which new technology and surveillance have historically ramped up worker production, but without raising wages.

She was, of course, absolutely right, and LMS’s are yet another, more comprehensive, version of this, one that—having divested itself of a range of library and administrative workers, while instituting new expectations of faculty—the vast majority of colleges and universities, as well as high schools and secondary schools, now depend on.

We can see a similar mission creep and transfer of labor in the implementation of LMS’s. The initial task of these platforms was to make course materials available to all students, regardless of how feckless, lazy, or confused they were, for free (and without incurring copyright lawsuits) wherever they were. Content included the syllabus as a downloadable document; readings in .pdf and on the electronic reserve at the library; and a design that broke the course down into week by week units. In the LMS’s, students would find readings, discussion sections, images, videos, class assignments (with grading rubrics and electronic deadlines that could be monitored), and explanations of why the students were doing what they were doing. Soon, LMS’s also developed grading and attendance widgets that drew on the evaluations we do throughout the semester. You could weight different assignments and evaluation criteria, press a button, and—presto! A numerically accurate grade popped up in seconds.

We should have known that this was a f*cked system from the get-go, of course. Why? Because early on, those wild and crazy experts in student life and academic affairs began to suggest—within five years of making LMS’s available—that students had better outcomes if you handed them a printed syllabus, and took notes by hand. What the LMS also meant was that every student, regardless of financial circumstances, had to have access to a computer. They could compete for one of the free terminals in the library and the various computer labs around the school—but that then made the course materials no more accessible, and papers no easier to hand in, than under the old reading and writing things on paper system.

OK? So that was a hint that this brave new world was not delivering good educational outcomes. And I haven’t even walked you through the number of days it takes to set one of these things up, even after you have a finished syllabus in hand. Nor will I address the thing no one wants to talk about: that these platforms have been sucking down student data for at least three decades, and either selling it or using it for proprietary development, without compensating anyone.

But like the Ginsu knife, there’s much, much more! By the turn of the 21st century, we entered a new era of metrics, standards, and scrutiny that mandated—wait for it—learning outcomes that had to be articulated and proven. In the case of high schools, learning outcomes that were testable were a sea change for curricula. At the college level, this mandate resulted in a drumbeat of requests for evaluations, self-studies, and reports that, because they avoided the standardization forced on secondary schools, were largely meaningless to the humanities and social sciences. Nevertheless, LMS’s were put to this task.

This too has created unwelcome chance. Before learning outcomes, a colleague of mine taught a first-year course about Henry David Thoreau’s Walden; Or, Life in the Woods (1854). Students were all instructed to purchase the book, which was the only reading in the course. The course description? “Together, we will read Walden.” The assignment every week? “Read Walden.” Which they did—the students were expected to read the whole book, every week, for eleven weeks, to deepen their insight into an American classic and the intellectual journeys it offered. It was a stunningly effective (and sneaky) way to, not just teach students to read at a college level, but also to get them interested in everything that 19th century United States history, literature, and philosophy have to offer. I would imagine that sometimes the learning outcome might be a profound sense that the 19th century, or history, or literature, or Henry David Thoreau, was to be avoided in the future at all costs.

Imagine this learning outcome on today’s syllabus: “Students will determine whether they do, or do not, wish to pursue future courses like this one;” or “students will determine whether they wish to spend the next four years reading difficult books, or doing problem sets in a STEM field.” Or, “students will fall in love with the American 19th century and decide to spend the rest of their lives there.”

LMS’s institutionalized stupid ideas about intellectual development through the demand for a rubric, a grid that tells students what they must do to get a certain grade, not what they might learn from a given assignment. “Rubrics,” media scholar Ian Bogost wrote in his own essay about being hamstrung during finals week by the Canvas outage,

are meant to avoid arbitrariness, but they also serve other instrumental goals: normalizing “learning objectives” so that universities can assess “learning outcomes” for accreditation and other bureaucratic purposes. This, in part, justifies the use of software such as Canvas, which allows instructors to write rubrics and grade against them, and (in theory at least) for assessors to roll up such results into reports and data. My assignment existed only inside Canvas, and my rubric along with it. I could not log in to see my own grading criteria and thereby offer my student advice about how to maximize the seven hours remaining until the assignment was due.

Paradoxically, rubrics take some of the teaching out of teaching, and at the same time, create more work for everyone. Just try grading a paper by locating and ticking off the nine to 15 squares of your rubric. Imagine being the student who is ticking off the same list as they prepare the paper for submission.

Rubrics are dismal for everyone, cementing in the minds of students that the final grade is the only point of the assignment. For faculty, evaluating (as opposed to reading) student work becomes a grim, rote activity that could easily (and I am sure, will) be turned over to an AI. Absent a rubric, however, you might realize that the student has completely misunderstood the question in an interesting way, or gone off on a tangent you didn’t anticipate and learned something you never intended.

In a rubric, that’s a C; in life, it’s an A.

When rubrics and learning outcomes dictate what ought to be learned they tend to render irrelevant things students might have learned that they will not be asked to show mastery of. LMS’s standardize what is taught and what students are required to respond to in order to demonstrate that they have learned that thing. Today, learning outcomes for the same course on Walden I described above would look something like this: “Students will leave the course understanding the long history of American environmentalism; the role of transcendental thought in 19th c. antebellum social thought; and how to do a critical reading of a foundational text.”

In order to show they have mastered these things, students would complete a series of graded and ungraded papers, perhaps a reading journal, evaluated according to a rubric that asks for specific elements. Yet other interesting questions that the typical first year student might come up with go unasked and unanswered. Where did Thoreau get clean underwear? Did the people of Concord think Thoreau was weird? What were the similarities between Thoreau and the Unabomber? Where were the Native Americans who used to live at Walden? Was this the kind of thing they taught at Harvard before the Civil War?

I am not saying that LMS’s destroy learning, but in their fruitless desire to standardize and report on human intellect, they set priorities that make eccentric and creative thinking irrelevant to college work. And the only real reason to have them are a series of financial incentives that benefit neither students nor faculty, whose data and labor is then returned to Big Tech as profit.

You know what else LMS’s do? Force reading onto screens where it doesn’t belong. One of the great complaints of today’s professoriate is that students “don’t read like they used to.” That’s true, but part of the problem is that their attention spans have been utterly fragmented by the screens they have been on, including in school, since they began their educations. How do I know this? Because adults, college professors, tell me the exact same thing about themselves. You can observe this phenomenon if you are attending a graduation this spring. If there are faculty up on the stage, count how many are not flipping through their mobile devices.

Furthermore, in what world do we simultaneously have bales of research that getting off screens improves attention and critical thinking, creates an atmosphere conducive to good sleep, reduces anxiety and shame—and then, push people under the age of 21, and their teachers, back to doing all their reading and writing online?

Students see the contradictions. As the editors of the Brown Daily Herald wrote last February in an appeal to bring back the printed course pack (which is basically a big binder of usually stolen work, bus so are the .pdfs posted on the LMS.) “For many reading-intensive classes,” they wrote, “the primary mode professors use to provide readings is to post a slew of files on the course’s Canvas page. While students in courses that prohibit laptops can print out each reading, doing so is costly and impractical.”

Studying this way also sucks compared to, you know, actual reading:

Course readers offer a rare luxury in the age of technology: the chance to focus deeply on one task at a time. When students online shop, browse LinkedIn or skim the news alongside their notes, they create only the illusion of engagement. But this multitasking can kill the efficacy of our learning. Even when class materials are the only thing open, notifications and the urge to open another tab can pull our attention away from the work at hand. By having students read from a physical packet instead of from a screen, professors will find fewer distractions and more critical engagement in their classrooms.

The elephant in the rooms from the Canvas hack is this: students and teachers alike are currently dependent on, and imprisoned in, a system that does not meet the needs of students or the university at large. Colleges and universities should begin to distinguish themselves by abandoning it, and investing the millions they spend on these site licenses in people.

Political Junkie

Discussion about this post

Ready for more?