Librarians Are Building Defenses Against AI-Generated Content Flooding Collections
Librarians are resisting AI adoption at rates significantly higher than other publishing professionals, according to survey data from the Book Industry Study Group. About a third of library respondents said they don't use AI and have no plans to start-compared to 20% across the full publishing dataset.
The resistance isn't coming from people unfamiliar with the technology. Librarians who reject AI tools can explain large language models and retrieval-augmented generation. They've made a deliberate choice based on what they believe a library should be.
The Real Problem: Low-Quality Content Clogging Collections
Library staff are spending time identifying and removing AI-generated "slop" from their catalogs. Librarians report that their collections are increasingly filled with low-quality AI-generated work-directly undermining their core mission to provide reliable information.
About 34% of librarians surveyed described themselves as ethically opposed to AI use. That figure matters because it comes from people with 11+ years of experience, many working in institutions with 100+ employees.
Less than half of library respondents said their institution actively uses AI. Thirty-one percent said they were unsure, likely because larger organizations don't always make decisions visible to individual staff members. Most libraries have no formal AI policy at all.
Where AI Actually Helps Libraries
The University of Texas music collection offers a practical example of AI solving a real problem. Thousands of albums and recordings sat effectively invisible to users because they'd never been fully cataloged. Traditional cataloging would have required resources the library didn't have.
A team used AI to create stub records, each tagged with a confidence factor. Records where the AI was least certain got flagged for human review later. High-use materials were prioritized for full human cataloging. The AI did the work humans couldn't afford to do, while humans retained control over accuracy.
This is a mission question, not a literacy question: Can AI help librarians do what they exist to do?
The Peer Review Crisis
AI is breaking the system that determines what society considers credible. For decades, peer review worked because the rate of human knowledge production roughly matched human capacity to review it. AI is shattering that balance.
We can't use the same AI tools accelerating content creation to evaluate that content's credibility-the trust isn't there. The result threatens not just publishing workflows but the broader infrastructure by which society establishes what is true.
Authors are already seeing the effects. Page reads in Kindle Unlimited may be declining partly because readers encounter a glut of AI-assisted and AI-generated titles. When readers lose trust that unfamiliar books will be worthwhile, they stick with what they know instead of taking chances on something new.
Libraries as Gatekeepers of Credibility
Librarians may be moving toward a new role: arbiters of reality. They function as "trusted humans in the loop," people who can vouch for a source because they have it in their collection, know where it came from, and can examine it directly.
Librarians have maintained credibility even as institutional trust has declined across society. That credibility matters. When a student gets an answer from AI, they've been trained to be skeptical. When it comes from the library, they tend to believe it.
Teaching people to question everything they encounter hits a psychological ceiling. The cognitive burden of constant epistemic vigilance becomes unsustainable. What people need isn't more instruction on AI skepticism but relief from the exhaustion of doubting everything.
That's what libraries offer: a place where the curation work is already done, where someone with expertise has already decided what's worth trusting.
- For writers concerned about AI-generated content: Understand how Generative AI and LLM systems work so you can better position your own work in an increasingly crowded marketplace.
- For those evaluating information quality: Learn about AI in research and how credibility assessment is changing.
Your membership also unlocks:
