Are AI education manifestos transforming from declarations into something entirely new: living documents that map our uncertainties?
The Manifesto Impulse
Higher education has a curious response to AI anxiety: writing manifestos—lots of them.
“Generative artificial intelligence (AI) has stormed higher education,” declares Ella McPherson and Matei Candea’s recent manifesto, “at a time when we are all still recovering from the tragedies and demands of living and working in a pandemic.” The sense of institutional overwhelm is palpable.
These documents follow a telling pattern. They begin with bold declarations—”establishing foundational principles” and “ensuring ethical deployment”—but quickly dissolve into qualification and complexity. Aras Bozkurt and colleagues’ Open Praxis manifesto admits that it “may not lead to generalizable findings, provide an exhaustive understanding, or reach a fixed conclusion.” So much for manifesto certainty.
As Trump’s return signals AI acceleration and deregulation, these institutional pronouncements will likely only multiply. Yet the manifesto—traditionally a vehicle for radical clarity—seems to crumble upon contact with AI’s complexities. What does this rush to declare positions tell us about our moment? And what gets lost when we pretend to have certainty about technology that outpaces our ability to understand it?
The Projects and Their Patterns
The Open Praxis manifesto, led by Aras Bozkurt with forty-six co-authors, reveals this tension most clearly. It begins with traditional manifesto ambition, seeking to “critically examine the unfolding integration of Generative AI.” But it quickly turns to metaphor, cataloging how we describe AI: “copilot,” “sorcerer’s apprentice,” “demon,” “bullshit generator,” “colonizing loudspeaker,” “stochastic parrot.”
When direct description fails, we reach for comparison. More telling still is how the document acknowledges its limitations. The authors note that “some concepts are intertwined and difficult to separate with sharp boundaries.” They admit that “due to the nature of the methodology, positive and negative aspects may inherently contradict each other.” This isn’t failure—it’s honesty about AI’s rapidly evolving nature.
McPherson and Candea’s manifesto proves especially revealing in its contradictions. While lamenting that GenAI arrived “without significant guidance,” it struggles to provide that guidance. Instead, it offers something more valuable: a framework for thinking about what we might lose. The authors worry about the “eureka moments” of scholarship—”the satisfaction of working out an argument through writing it out, the thrill of a sentence that describes the empirical world just so, the nerdy pride of wordplay.” Their admission that “ethical frameworks are racing to catch up with research practices on new terrains” is even more striking. They advise following “internet researchers: follow your instinct (if it feels wrong, it possibly is) and discuss, deliberate and debate.” This retreat to gut feeling and collective discussion speaks volumes.
The Safe AI manifesto, authored by Marc Alier Forment, Francisco Garcia Peñalvo, and colleagues, takes perhaps the most practical approach. It offers seven principles for deploying AI.
Yet its most notable feature is structural: It’s designed as a living document, openly acknowledging that any guidance offered today might need revision tomorrow. “This manifesto will be updated,” they write, “as the community and the technology mature.” This admission—that today’s certainties might not be tomorrow’s—feels remarkably clear-eyed.
Margarida Romero and colleagues’ “Human-Centered Education” manifesto attempts to split the difference between principle and practice. It introduces the concept of “hybrid intelligence” and proposes a six-level model for AI engagement in education, from passive consumption to “expansive learning.”
Yet even here, complexity dominates. The authors acknowledge how AI simultaneously “broadens access to information” while “exacerbating digital divides” and might “streamline tasks” while generating “additional work through thorough fact-checking.” Rather than resolve these tensions, they present them as inherent to our moment. They note that “interventions in one part of the AI ecosystem (e.g., the need for learners’ privacy) can have consequences in other parts (e.g., using facial recognition to identify the learners’ engagements).”
The Value of Uncertainty
What emerges isn’t a path forward but a map of our uncertainties. Each manifesto reveals the challenge of writing about technology that’s actively reshaping how we write. Their value lies not in their declarations but in their documentation of this struggle. That’s precisely what we need—not confident pronouncements about AI’s place in education but honest wrestling with its complexities.
The manifesto impulse is understandable. As AI regulation retreats and institutional pressures mount, we want certainty—clear principles, firm guidelines, and solid ground. But these documents suggest a different approach. Instead of racing to certainty, we should embrace the productive discomfort of this moment. The Safe AI manifesto’s “living document” approach points the way: our frameworks must evolve as rapidly as the technology they address.
This might mean reimagining what a manifesto can be. It is not a declaration of unchanging principles but a document that grows with our understanding. It is not a solution to uncertainty but a framework for engaging with it. Most importantly, we must acknowledge that our relationship with AI—in education and beyond—will be marked by constant evolution and necessary revision.
The manifesto moment reveals something crucial: our institutional responses to AI often say more about our anxieties than our understanding. We’ve documented our confusion in trying to write our way to clarity. That’s not a failure. It might be the most honest starting point for whatever comes next.
About the author
Dr. Nick Potkalitsky is transforming education through AI integration, leveraging his experience teaching across diverse settings from The Miami Valley School to Ohio State University, where he served as a Graduate Teaching Associate. His work spans language arts, history, media studies, and digital humanities.
As CEO of Pragmatic AI Solutions and publisher of Educating AI on Substack, he shares practical insights through curriculum briefs and research summaries. His expertise has earned him speaking engagements at notable events including the Chicago Book Fair, Edu-Con, and the Ohio Educational Technology Conference.
With degrees from The Ohio State University, John Carroll University, Cleveland State University, and Oberlin College, Dr. Potkalitsky brings academic rigor to educational innovation. Based in Dayton, OH, he continues to shape AI literacy in education through his teaching, writing, and professional training programs.
This article has been republished with permission from Nick Potkalitsky, PhD, from the LinkedIn group The Pragmatic AI Educator.
