The One Rule
AI in Libraries and Human Flourishing
Infophilia: A Positive Psychology of Information | May 9, 2026 | Vol. 4, Issue 25 | Choose Privacy Week — Open Access Edition | Updated Notes: 05/09/26, 9:43 am
Artificial Intelligence | Infophilic Information Styles | Intellectual Freedom | Library and Information Science | Technophilia | Wellbeing
✨ Welcome to Infophilia, a weekly serial and a living lab exploring how our love of information and connection can help us thrive, individually and collectively. This is avant-garde research in how we engage with knowledge, meaning, and each other.
📌 Access & Attribution: This serial offers free previews always and occasional open access essays to keep scholarship accessible. Students and those facing financial barriers can request complimentary access. If you find value here, please cite the original work and consider supporting the public scholarship through subscription. Proper attribution sustains this work and models healthy information engagement.
Cite as: Coleman, A. (2026). The one rule: AI in libraries and human flourishing. Infophilia: A Positive Psychology of Information 4(25).
The One Rule
AI in Libraries and Human Flourishing
I promised you a lighthearted and fun spring of readings. This week I want to make good on at least part of that promise before we get to the heavier sections below.
Thanks to readers who wrote in after last week’s What the Teapots Know. Using my own good news / bad news framing, except this time it is all good news, I share one below (others in due time) — thank you, Mrs. M!
First: Fix the News (link in Notes) by Angus Hervey. If you need evidence that the world is also producing discoveries, breakthroughs, and extraordinary things to wonder alongside everything else, this is it. It is not toxic positivity. Evidence. Read it. 😊
Second: Mrs. Hummingbird.
We recently discovered her nest among the geranium branches spilling over from a hanging basket on our porch. The basket sways with the wind and the branches too. When I began preparing this essay last week she was sitting on two eggs; the eggs hatched Friday, and voila, we’ve a new mother in time for Mother’s Day this weekend.
Our feature image today is her, a small, determined little bird, building something fragile, in a world that keeps moving. The 2 min video here is a candid camera of a new mom.
Happy Mother’s Day to all the moms out there taking care of the curious.
Now for today’s essay: two field notes from the front line.
Authentication & Access
Why Firefox
What the ALA Core Authentication and Authorization Committee’s notice about impending browser technology changes really means, and why Chrome and Safari are not the answer
The web is getting more private.
I recently paid for another year of Wired, couldn’t access a single issue, emailed customer service, got bounced to support, sent my invoice ID and customer number, and heard nothing for a week. I wrote a cancellation letter. Within hours, I received a cheerful reply commiserating with my frustration and explaining that, per policy, they couldn’t refund me. But I was welcome to access the magazine. There was no mention whatsoever of the fact that I couldn’t access it.
Losing access is not just a Wired problem. This is a browser problem. It is coming for everyone.
Here’s what’s actually happening. For years, the web relied on little invisible handshakes (cookies, tracking tokens, silent redirects) to keep you logged in everywhere and remember who you were. Publishers, libraries, and subscription services built access systems on top of these handshakes.
Now Firefox, Safari, and (slowly, reluctantly) Chrome are blocking all of that. For privacy reasons. This is good news: you are being tracked less. The trade-off is that authentication systems that were never fully modernized are quietly breaking. And when they break, support teams often can’t see the problem because it isn’t in their system. It is in the missing handshake.
The American Library Association (ALA) Core Authentication and Authorization Committee put out a detailed update on all this recently. It was written for systems librarians and was, let’s say, bracing. So, here’s the short version for the rest of us:
What’s changing:
Third-party cookies are going away. This is why “stay logged in” stops working, why your institution isn’t remembered, and why global logout has become unreliable.
Bounce tracking is being blocked. Redirecting you through third-party sites to collect data no longer works, which also means some authentication flows break.
IP-based access is wobbling. Safari now limits IP visibility by default. IP authentication still mostly works; but “mostly” is doing a lot of work there.
None of this tells you what to actually do. So here’s that:
More frequent logins are common now. This is not a bug, not your fault, and it is not fixable by clearing your cache one more time (although it can sometimes help).
When something breaks, try a private window and log in fresh from the publisher’s homepage (not a saved bookmark or email link).
Check which email or login method you used. Many systems separate billing identity from login identity.
Prefer “sign in with your institution” over IP access where possible.
In Safari, your searches may still go to Google even though the browser blocks trackers, because Apple is paid billions per year to keep it that way. If that bothers you, change your default search engine (e.g. DuckDuckGo).
Why Firefox? Because Firefox is the only major browser that blocks third-party cookies and bounce tracking by default, without asking you to opt in, without a business model that depends on your data. Chrome is slowly moving in this direction while protecting Google’s advertising revenue. Safari blocks some things but still lets Apple collect what it needs.
The systems are catching up. In the meantime, expect more logins. And if support replies empathize warmly without addressing the actual issue, that’s not a privacy feature. That’s just customer service.
Call it what you will. 😊
AI in Libraries and Human Flourishing
The One Rule
Ethical augmentation and what AI actually does well
The ALA Artificial Intelligence Policy Working Group has released a draft guidance document for libraries. This is a good document. It cites Luciano Floridi. It names intellectual freedom, public good, privacy, sustainability, equity. It says, in one quiet clause buried in its appendix, that AI should enable “human flourishing — including mental and social wellness.”
Human flourishing. The guidance document named it.
But neither it nor we may know what that means yet. That’s what this section is for.
But before we delve into flourishing, we need the Rule.
One Rule. Not a hundred prompt engineering tricks.
I once asked Claude directly about its tendencies to sycophancy, lying, and to generate confident, fluent claims that are sometimes false, sometimes partially true, sometimes true but distorted. The answer was instructive (emphasis is mine):
“I make things up. Not randomly, and not maliciously, but I confabulate. I’m a language model trained to produce coherent, fluent text, and coherence can be achieved with false content just as easily as with true content. I don’t have reliable access to the difference from the inside. Fluency masks uncertainty. I sound equally confident when I’m on solid ground and when I’m not. That’s a genuine danger. The corrective isn’t to distrust everything — it’s to verify anything that matters, especially anything that carries your name.”
From that answer, I crafted one Rule.
Question the chatbot back on everything. Verify anything that carries your name. Triangulate, with other AI tools and with non-AI sources.
Researchers used to call the third motion triangulation. Research now shows that the most sophisticated AI users are doing it instinctively: multiple chatbots, cross-checked against primary sources, experts, libraries, and lived experience. They haven’t abandoned critical thinking. They’ve augmented it.
The AI Rule is not for academics or writers or researchers or students alone. It is for everyone: the patron at the reference desk, the student writing their first paper, the retiree researching a health diagnosis, the school librarian in a Title I school trying to serve students with a limited budget and a generous heart, the information professional trying to make sense of a landscape changing faster than thought itself.
Question it back. Verify. Especially before you use or adapt what carries your name. Use multiple sources, including ones that were made by humans who cared (hint: library workers of all stripes, zine-creators, journalists, and computing wizards.)
The profession’s quiet struggle
Librarians know this better than almost anyone. They struggle with it more than they say publicly.
Across the community of practitioners writing about AI and libraries, a pattern is visible: smart, generous professionals building tool lists, translating policy documents, designing programs, trying to help patrons navigate an extractive landscape while simultaneously navigating it themselves. Hurt, bewildered people trying to get control. Making a living. Helping.
One practitioner named it directly: the cognitive dissonance between the widespread exhortations to adopt GenAI tools in libraries and the harms we see. Most don’t say it that plainly.
As Bill Badke put it this month in his InfoLit Land column, librarians need to recognize that building creativity in robots is a frequent goal of our work: helping students break out and be humans rather than robots. That is also the infophile’s challenge: not to become the robotic researcher who follows steps without growing, delivers products without becoming, and reaches for AI because the research project is a tribulation rather than a quest.
What is rarely named is the deepest threat: technology is now changing faster than thought itself, while the profession historically responsible for preserving the conditions for slow, reflective thinking has not fully developed the language for what it is protecting.
Libraries have always been more than access points and retrieval systems. They have also been places where people construct the future selves who will read, integrate, question, and act on what they find. The library is not a church, but it has long maintained traditions concerned with the inner life of information encounter: bibliotherapy, contemplative space scholarship, slow librarianship, developmental reading practices. These traditions exist, but largely in isolation from one another and from mainstream AI policy discussions.
Nothing new under the sun
There is a harder question. The one the Rule doesn’t fully answer: what was AI built on?
I headlined it myself: “When You Chat with AI, It’s Using 192,640 ‘Borrowed’ Books — And That’s Just the Start.” The Anthropic settlement. HathiTrust and Internet Archive restrictions. The Library of Congress’s By the People volunteers whose transcription labor became training data while downstream value flowed to those with capital and scale.
We’ve been little thieves all along. Q’ohelet got it right: there is nothing new under the sun. Every knowledge system is built on what came before it, sometimes with permission, sometimes without. The oral poets who borrowed epics. The scribes who copied manuscripts. The search engines that indexed the web. The language models that trained on our books, our essays, our emails, our lives, our dissertations.
What do we owe the builders? And what are we building toward? (Q'ohelet had something to say about this too — a thread for another week.)
For now, here's what the extraction model looks like at its most visible.
Coral Hart and the infofool end
Coral Hart, the UK’s romance queen, is now writing 200 AI-generated novels a year and teaching others to do the same.
This is the infofool end of the spectrum dressed up as productivity. Technically proficient. Industrialized creativity. Adaptive information extraction masquerading as creativity.
Hart isn’t doing anything entirely new. The publishing industry has long rewarded formula and scale. AI simply accelerates the extraction model: more content, less friction, thinner meaning. The infophile’s question is different here. It is not: How much can I produce? It is: “What relationship to knowledge and creation am I cultivating while I produce it?”
The infophile produces differently. The infophile is changed by what they make.
What AI actually does well: the Socratic mirror and the panchayat model
Here is what I have never written publicly.
AI hasn’t made me smarter. It has made me clearer.
When I use an LLM chatbot in Socratic dialogue — pushing back, questioning its assertions, demanding meticulousness, refusing its first fluent answer — it reflects my own thinking back to me with a clarity I couldn’t achieve alone. I can see my assumptions. I can locate my biases. I can name where my expertise ends and my wishful thinking begins. Not because the AI is wise. Because the mirror is clean. Almost. (What it reflects is another important question, for another week.)
Most people use AI as an oracle: they ask questions and accept the answers. But oracles are dangerous exactly because their answers sound authoritative. The Socratic use is different: to discover what you already half-know or know you don’t know but cannot yet articulate.
The second move is the panchayat model. In South Asian democratic tradition, the panchayat is the village council. There is collective sense-making: wisdom arriving through dialogue rather than individual authority. Like all councils, panchayats can also reproduce gatekeeping and hierarchy. I use chatbots the same way: not as experts but as councils. Different models, different framings, different provocations. The disagreements between them are as useful as the agreements. The gaps in their knowledge tell me where the real questions live. And I am never too busy to ask: what is missing?
The Socratic mirror plus panchayat precision, together they describe an infophile’s relationship with AI: augmentation of the capacity for self-knowledge and collective sense-making.
AI, much like a human librarian, mediates the flow of information. Used with foresight, it can deepen curiosity, reflection, and collective sense-making while still preserving what makes information meaningful and ethically grounded. But only if the infostyle we bring to the encounter is authentically communal rather than extractive.
The Tamil Sangam poets understood this. Knowledge was not possessed; it was made in the space between minds, tested against the community, refined through response. And whether you believe in a Divine Creator or not, the Bhakti tradition went further: the devotee’s relationship to the divine was itself a form of knowing: relational, cultivated, practiced. Arivu anbu aram: knowledge, love, ethics, inseparable.
AI can be a Sangam partner or a void-filler, depending entirely on how you approach it. The difference is not in the technology. It is in the infostyle you bring to the encounter.
What the ALA document almost says
The ALA AI guidance draft, buried in its appendix, says AI should enable “human flourishing — including mental and social wellness.” That clause, one of the most important sentences in the document, is the least developed.
What does flourishing look like? The document doesn’t say. It has a rich vocabulary for harm: misinformation, cognitive offloading, algorithmic bias, data extraction. It has almost no vocabulary for its opposite. It can tell you what libraries are protecting patrons from. It cannot yet tell you what it’s cultivating toward.
That vocabulary exists. Kate Langan's framework for flourishing with information, organized around belonging, confidence, and connectedness, offers one such cultivation. Thirty years of information literacy scholarship have not yet been synthesized into a shared framework for flourishing in AI-mediated environments.
Emerging scholarship increasingly argues that AI ethics should be framed around human flourishing rather than harm reduction alone.
This is one attempt at synthesis. Floridi describes the infosphere as a commons to be actively stewarded. The infophile seeks, connects, creates meaning, and returns value to the commons they draw from. The difference between infofool and infophile is a matter of practice and orientation.
The Rule is simple: question everything, verify what carries your name, triangulate across sources. Because you are responsible for what you make with it.
Ethical augmentation is not a compromise between using AI and refusing it. It is a third thing: using AI in the service of your deepest information drives — curiosity, connection, meaning-making, the public good — rather than letting it exploit your vulnerabilities or replace your thinking.
That is what the ALA document means when it says human flourishing. It just needs the vocabulary to fill it in. Adaptive infophilia offers one way to name and practice that work.
Closing: The Question We’re Left With
This Choose Privacy Week we have covered the technical face of the privacy struggle — browsers, cookies, authentication, the invisible handshakes that are quietly breaking — and the philosophical face: what it means to engage with AI as an infophile rather than an infofool.
But there is a third face neither document fully addresses.
If AI is now shaping not just what we find but what we think to ask, then the real task of ethical augmentation is preserving the future self who will still want to know.
What does it look like when we succeed at that?
Mrs. Hummingbird knows. She built something fragile in a world that keeps moving, and she sits on it anyway.
The library, at its best, has always been that answer.
Notes
This essay responds to two ALA documents: the ALA Core Authentication and Authorization Committee’s update on browser technology changes (2026) and the ALA Artificial Intelligence Policy Working Group’s Draft Guidance on the Use of AI in Libraries (April 2026), available at ala.org. The ALA AI guidance document's theoretical apparatus cites Floridi and one dictionary entry, but no practitioner literature. For a document addressed to library practitioners, that absence is worth noting. The author submitted formal feedback to the AI working group on May 5, 2026; the full feedback is available on request. [05/06/26 10 am addition: The feedback also mentioned Gorman’s enduring values. Gorman, M. (2000). Our enduring values: Librarianship in the 21st century. ALA Editions. See also: Gorman, M. (2015). Our enduring values revisited: Librarianship in an ever-changing world. ALA Editions. These foundational statements map directly onto the ALA AI guidance draft’s values vocabulary, though the draft doesn't name them. Gorman identifies eight core values of librarianship (stewardship, service, intellectual freedom, privacy, rationalism, literacy and learning, equity of access, and democracy) grounded in liberal, democratic, and humanist principles, and deeply influenced by S.R. Ranganathan, whom Gorman regarded as "the greatest figure of librarianship in the 20th century." Gorman's framework has been criticized for insufficient attention to diversity, decolonization, and Global South LIS; but the guidance document doesn't make that mistake and is attentive to sustainability as well.]
(A third document, the revised Framework for Information Literacy for Higher Education was also open for feedback during the same time.)
The theory underlying this essay is developed in: Coleman, A.S. (2024). An introduction to infophilia, a positive psychology of information. Library Research Seminar 8: Telling Library Stories. https://uknowledge.uky.edu/lrs8/3/ | It was also published as a peer-reviewed paper in the Indian journal SRELS International Journal of Information and Knowledge, 62(1): https://doi.org/10.17821/srels/2025/v62i1/171701 | For information weaponization see Coleman, A.S. (2026). Adaptive infophilia: The ethics of information power [Preprint] SocArXiv. https://doi.org/10.31235/osf.io/kjh6w_v1 | The “future self” referenced in the closing is available in the full open access essay on self-censorship: https://infophilia.substack.com/p/the-self-censorship-they-didnt-want
On Bill Badke and InfoLit Land
Badke has been writing about AI and information literacy in his InfoLit Land column (Online Searcher / Computers in Libraries) since 2008. Key columns include:
Badke, W. (2015). The effect of artificial intelligence on the future of information literacy. Online Searcher, 39(4), 71-73.
Badke, W. (2023). AI challenges to information literacy. Computers in Libraries, 43(3), 41-42. Reprinted in Information Today Europe, January 2, 2024. https://www.infotoday.com/cilmag/apr23/Badke--AI-Challenges-to-Information-Literacy.shtml
Badke, W. (2025). The great AI rubbish heap. Computers in Libraries, 45(4), 39-40. https://www.infotoday.com/cilmag/may25/Badke--The-Great-AI-Rubbish-Heap.shtml
Badke, W. (2026). The robotic researcher. Computers in Libraries, 46(3), 39-40. https://www.infotoday.com/cilmag/may26/Badke--The-Robotic-Researcher.shtml
None of these columns appear in the ALA guidance draft’s theoretical apparatus.
On the profession’s struggle with AI
Coffey, L. (2024). Librarians want to adopt AI but cite lack of expertise. Inside Higher Ed. https://www.insidehighered.com/news/quick-takes/2024/09/10/ai-adoption-top-concern-librarians-next-year
Fox, V. (2025). AI refusal in libraries: A starter guide. ACRLog. https://acrlog.org/2025/06/11/ai-refusal-in-libraries-a-starter-guide/
Gültekin, V., & Kavak, A. (2025). An assessment of artificial intelligence anxieties of academic librarians. Journal of Academic Librarianship.
Luo, L. (2025). Use of generative AI in aiding daily professional tasks: A survey of librarians’ experiences. Library Trends. https://muse.jhu.edu/article/961200
Paul, A. (2025). Librarians can’t keep up with bad AI. Popular Science. https://www.popsci.com/technology/librarians-bad-ai/
Tanzi, N. (2025). AI presents a communication challenge for libraries. The Digital Librarian. https://the-digital-librarian.com/2025/12/04/ai-presents-a-communication-challenge-for-libraries/
On AI literacy and digital wellbeing
Clarivate. (2025). Pulse of the Library 2025. https://clarivate.com/pulse-of-the-library/
Kaya, B., Çınar, S., & Cenkseven Önder, F. (2025). AI Literacy and Digital Wellbeing: The Multiple Mediating Roles of Positive Attitudes Towards AI and Satisfying Basic Psychological Needs. International Journal of Human–Computer Interaction, 1–11. https://doi.org/10.1080/10447318.2025.2499162 — AI literacy is positively associated with digital wellbeing, mediated by positive attitudes toward AI and the use of AI for satisfying basic psychological needs.
On AI and human flourishing
Knowles, M.A. (2021). Five motivating concerns for AI ethics instruction. Proceedings of the Association for Information Science & Technology, 58(1). https://doi.org/10.1002/pra2.481
Langan, K. (2025). A framework for flourishing with information: Centering the value of academic libraries on student belonging, confidence, and connectedness. ACRL/ALA. https://www.ala.org/sites/default/files/2025-03/AFrameworkforFlourishingwithInformation.pdf — Langan also maintains a LibGuide at Western Michigan University.
Poole, A.H. (2023). Data flourishing: Developing human-centered data science through communities of ethical practice. Proceedings of the Association for Information Science & Technology, 60(1). https://doi.org/10.1002/pra2.793 — The 2023 ASIS&T Annual Meeting theme, “Making a difference: Translating Information Research into Practice, Policy, and Action,” framed its aspirations around “promoting the power of information to develop human happiness, equality, and wellbeing.”
VanderWeele, T.J., & Teubner, J.D. (2026). Flourishing considerations for AI. Information, 17(1), 88. https://doi.org/10.3390/info17010088
On good news and the world
Hervey, A. Fix the News.
Something to look out for.
--Magnifica Humanitas, on the safeguarding of the human person in the age of AI — will be presented at 11:30am on Monday, May 25, in the Vaticanʼs Synod Hall, in the presence of the Holy Father. source: https://www.vaticannews.va/en/pope/news/2026-05/pope-leo-xiv-first-encyclical-magnifica-humanitas.html