What happens when your employer turns your expertise into a product—and doesn't bother telling you? That's the question Arizona State University has forced its faculty to answer. The institution recently launched "Atomic," an AI platform that chops professor lectures into short clips and generates learning modules from them. Chris Hanlon, an ASU professor of US literature, discovered his own face staring back at him from an AI-generated module. He hadn't been notified. None of his colleagues had either.
The revelation has sparked fury among ASU faculty, and the incident reveals something far more troubling than a botched product launch. Atomic isn't a failure—it's a blueprint.
Faculty learned about Atomic through word of mouth, not official communication. Hanlon found a one-minute clip extracted from a twelve-minute lecture, with an AI transcription mangling the name of literary critic Cleanth Brooks into "Client Brooks." The context that gave that moment meaning—surrounding discussion, student questions, the semester's arc—vanished entirely. What remained was slop dressed in academic packaging.
404 Media tested the platform and confirmed the academic weakness. More striking: anyone could sign up for a free twelve-day trial with a personal email address. No ASU affiliation required. The platform was selling access to faculty expertise to the general public, and the faculty whose labor made it possible weren't asked, weren't told, and weren't compensated.
ASU claims Atomic targets alumni and "those who previously expressed interest in ASU's learning initiatives." But the mechanism is clear: take existing lectures, feed them to AI, generate derivative products, monetize them. The faculty are the raw material. The university keeps the profit.
Universities have always owned intellectual property created during employment. But this represents something different—a systematic industrialization of faculty expertise without consent, consultation, or concern for academic integrity. The AI-generated modules don't just strip context; they actively produce inaccuracies that faculty would never endorse.
Hanlon's colleagues whose lectures appeared in that module were all equally blindsided. The pattern suggests this wasn't an oversight but a design choice. Notify faculty, and they might object. Launch quietly, and the platform becomes fait accompli before anyone objects.
The academic community is watching. Other universities are almost certainly developing similar systems. Atomic shows them exactly how to do it: extract value from existing labor, minimize faculty agency, deploy AI as both the mechanism and the justification, and frame everything as "improving the learner experience."
Hanlon's one-minute clip about Cleanth Brooks now exists in a form he never sanctioned, accuracy mangled, context stripped, monetization flowing elsewhere. He learned about it from social media. That's the real lesson of Atomic—not what happens when AI fails, but what happens when it works exactly as designed.