Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction
Open accessThe DOI matches its Crossref record; 26 of the 26 references it cites were found in a registry or an index; none of the 16 sources its claims cite is checked for support yet.
Abstract
As librarians grapple with decisions over when and how to incorporate generative artificial intelligence (AI) into information literacy instruction, they must also consider the barriers users face in accessing these tools. This article will examine generative AI through the lens of information privilege to better understand how librarians can prepare to teach students about the limitations they could face while using it both in and out of higher education institutions. An analysis of eleven AIs designed to assist with literature reviews reveals that the “freemium” model, where users of the free version do not have access to features available in a paid subscription, currently dominates the emerging AI research assistant market. Additionally, to make full use of these tools, users must already have a high level of information privilege due to their reliance on users supplying or being able to access paywalled content. Implications for including generative AI tools in information literacy instruction are discussed, including suggestions for how these tools can be leveraged to introduce the concept of information privilege.
Overview
CoChat Summary
TL;DR
This article analyzes eleven generative AI research assistants through the lens of information privilege, finding that most use a freemium model that limits free users’ features or usage and often requires preexisting access to paywalled sources. It argues librarians should account for these access barriers when incorporating such tools into information literacy instruction and can use the tools’ limitations to teach information privilege.
Key findings
- Of eleven AI research assistants analyzed, one was free (Semantic Scholar), one required a paid subscription (scite), and eight used an explicit freemium model; Paper Digest allowed limited anonymous use but required a subscription to create an account.
- Among freemium tools, free versions most commonly restricted how often the tool could be used, restricted features, or both; two limit models emerged: daily/monthly use limits that reset, and limits on how many sources or how much of a source a user can interact with.
- Full use of many tools depends on users already having information privilege: tools with discovery features still lead to paywalled results, and tools requiring uploaded PDFs or full text require preexisting access to sources; some tools also put advanced AI models (e.g., GPT-4) behind subscriptions.
- For instruction, the author recommends prioritizing forever-free/no-account tools, demonstrating account-required tools from the librarian’s account, providing alternatives for students who opt out, and pairing instruction with information privilege/open access education, including preprints and paywall distinctions.
Methods
- The study analyzed eleven generative AI research assistants selected because their primary functions included at least one literature-review task (source discovery, identifying relevant information, summarization, or synthesis) and because they generated original text; ResearchRabbit, Litmaps, and Connected Papers were excluded because they did not generate original text at the time of the study in spring 2024.
- Tools were assessed for free versions, differences between free and subscription versions, institutional subscription options, and discounted pricing, using pricing/FAQ pages and testing of primary generative functions.
- For tools with discovery features, the same two searches—“feline diabetes mellitus” and “How are feline diabetes mellitus and feline pancreatitis related?”—were entered; initial analysis occurred April–May 2024 and was confirmed/updated August 2024, with paywalled search features tested via free trial.
Limitations
- The paper states that the tool descriptions are not meant to be reviews or exhaustive overviews, and that the generative AI landscape is rapidly evolving, requiring regular review of pricing and restrictions.
- The paper states that whether generative AI provides a true advantage remains to be seen and that comparing completeness of traditional database searches with AI research assistant searches may prove difficult due to contrasting search strategies/prompts.
- Evident from the text: the analysis is limited to eleven tools and a specific spring/August 2024 snapshot; it tests free-account/free-trial experiences rather than subscriber experiences, and it does not include user studies or student outcomes.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 17 Not valid · 13 Not checked · 84 Own results and interpretations
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, four tools put one feature of particular note behind their paywalls: the advanced version of the AI model powering their generative features.”
- Also stated in: § Discussion · plain
“four tools put one feature of particular note behind their paywalls: the advanced version of the AI model powering their generative features.”
§ Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § Discussion · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“As described in the analysis, many generative features either do not work or produce minimal output when not supplied with full text.”
- Also stated in: § Discussion · plain
“many generative features either do not work or produce minimal output when not supplied with full text.”
§ Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction > Discussion
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § Discussion · plain
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Of the ten subscription tools, seven at the time offered options for institutional subscriptions, a strong indication that their developers seek to capitalize on the same market that publishers and database vendors have long found so profitable.”
- Also stated in: § Discussion · plain
“Of the ten subscription tools, seven at the time offered options for institutional subscriptions”
§ Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction > Discussion
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: unclear — It is unclear why the paper says it holds.
- Also stated in: § Discussion · plain
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“There is no visual distinction between open access and paywalled search results.”
§ Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction > Additional information privilege implications
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“While users can filter results to open access articles, there is no visual indicator of which results are open access and which are paywalled in the list of results or item record.”
§ Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction > Additional information privilege implications
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
26 more claims
Create a free account to see all 114 claims and to verify the 16 references still open.
- It may state two separate claims as one. 35
- Probably not a substantial claim. 77
- Borderline: one of the tests for a substantial claim is unclear. 1
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction > Discussion
“two tools with search functionality (Elicit and Paper Digest) allow unlimited searches and only restrict using the generative AI to summarize and paraphrase.”
§ Generative AI and the Reinforcement of Information Privilege: Implications for Information Literacy Instruction > Usage restrictions
“Free account users can perform unlimited searches but cannot export their spreadsheets.”
Citations
Cites
Students’ voices on generative AI: perceptions, benefits, and challenges in higher education
Chan, C. K. Y., & Hu, W. (2023) · International Journal of Educational Technology in Higher Education · doi:10.1186/s41239-023-00411-8 · Cited by 2,301 · cited 1×
“…the benefit of tools that could help them locate and interpret sources for research projects, they also express concerns about the risk of becoming too reliant on the technology (Aure and Cuenca 2024; Chan and Hu 2023) and about the reliability of generative AI outputs (Deschenes and McMahon 2024).”
ChatGPT: five priorities for research
van Dis, E. A. M., Bollen, J., Zuidema, W., van Rooij, R., & Bockting, C. L. (2023) · Nature · doi:10.1038/d41586-023-00288-7 · Cited by 1,842
Zhai, C., Wibowo, S., & Li, L. D. (2024) · Smart Learning Environments · doi:10.1186/s40561-024-00316-7 · Cited by 1,567
Generative AI in Education and Research: Opportunities, Concerns, and Solutions
Alasadi, E. A., & Baiz, C. R. (2023) · Journal of Chemical Education · doi:10.1021/acs.jchemed.3c00323 · Cited by 463
Generative AI: Here to stay, but for good?
Sætra, H. S. (2023) · Technology in Society · doi:10.1016/j.techsoc.2023.102372 · Cited by 312
The ChatGPT Storm and What Faculty Can Do
Sun, G. H., & Hoelscher, S. H. (2023) · Nurse Educator · doi:10.1097/nne.0000000000001390 · Cited by 252
Nicholson, J. M., Mordaunt, M., Lopez, P., Uppala, A., Rosati, D., Rodrigues, N. P., Grabitz, P., & Rife, S. C. (2021) · bioRxiv (published in Quantitative Science Studies) · doi:10.1101/2021.03.15.435418 · Cited by 202 · cited 1×
“Developers of the tool scite describe large language models’ potential to provide context on how an article was cited in another paper, rather than simply relying on citation metrics that show how often a source has been cited as a sign of quality (Nicholson et al. 2021).”
Qureshi, R., Shaughnessy, D., Gill, K. A. R., Robinson, K. A., Li, T., & Agai, E. (2023) · Systematic Reviews · doi:10.1186/s13643-023-02243-z · Cited by 201
Semrl, N., Feigl, S., Taumberger, N., Bracic, T., Fluhr, H., Blockeel, C., & Kollmann, M. (2023) · Human Reproduction · doi:10.1093/humrep/dead207 · Cited by 68
Cheng, S.-L., Tsai, S.-J., Bai, Y.-M., Ko, C.-H., Hsu, C.-W., Yang, F.-C., Tsai, C.-K., Tu, Y.-K., Yang, S.-N., Tseng, P.-T., et al. (2023) · Journal of Medical Internet Research · doi:10.2196/51229 · Cited by 50
A Survey on Student Use of Generative AI Chatbots for Academic Research
Deschenes, A., & McMahon, M. (2024) · Evidence Based Library and Information Practice · doi:10.18438/eblip30512 · Cited by 49 · cited 1×
“…the benefit of tools that could help them locate and interpret sources for research projects, they also express concerns about the risk of becoming too reliant on the technology (Aure and Cuenca 2024; Chan and Hu 2023) and about the reliability of generative AI outputs (Deschenes and McMahon 2024).”
Visual Exploration of Literature Using Connected Papers: A Practical Approach
Behera, P. K., Jain, S. J., & Kumar, A. (2023) · Issues in Science and Technology Librarianship · doi:10.29173/istl2760 · Cited by 48
Aure, P. A., & Cuenca, O. (2024) · Journal of Research in Innovative Teaching & Learning · doi:10.1108/jrit-03-2024-0076 · Cited by 42 · cited 1×
“…the benefit of tools that could help them locate and interpret sources for research projects, they also express concerns about the risk of becoming too reliant on the technology (Aure and Cuenca 2024; Chan and Hu 2023) and about the reliability of generative AI outputs (Deschenes and McMahon 2024).”
James, A., & Hampton Filgo, E. (2023) · College & Research Libraries News · doi:10.5860/crln.84.9.334 · Cited by 41
Visual citation navigation of open education resources using Litmaps
Kaur, A., Gulati, S., Sharma, R., Sinhababu, A., & Chakravarty, R. (2022) · Library Hi Tech News · doi:10.1108/lhtn-01-2022-0012 · Cited by 37
Warren, S. A., & Duckett, K. (2010) · Journal of Library Administration · doi:10.1080/01930821003667021 · Cited by 35
Bissessar, C. (2026) · Equity in Education & Society · doi:10.1177/27526461231215083 · Cited by 34
Literature review in the generative AI era - how to make a compelling contribution
Pan, S. L., Nishant, R., Tuunanen, T., & Nah, F. F. (2023) · The Journal of Strategic Information Systems · doi:10.1016/j.jsis.2023.101788 · Cited by 29
Pedagogy and the Logic of Platforms
Gilliard, C. (2020) · Cited by 25
Hall, T. D. (2021) · Journal of Library Administration · doi:10.1080/01930826.2021.1906559 · Cited by 21
(2017) · College & Research Libraries · doi:10.5860/crl · Cited by 19
Güner, H., Er, E., Akçapınar, G., & Khalil, M. (2024) · DOAJ (DOAJ: Directory of Open Access Journals) · doi:10.30191/ets.202404_27(2).tp05 · Cited by 12
Evaluating the Accuracy of scite, a Smart Citation Index
Bakker, C., Theis-Mahon, N., & Brown, S. J. (2023) · Hypothesis: Research Journal for Health Information Professionals · doi:10.18060/26528 · Cited by 10
Can generative AI facilitate the research process? It’s complicated
Kingsley, D. (2023) · College & Research Libraries News · doi:10.5860/crln.84.9.342 · Cited by 6
Open Access for Scholars Left Behind: The Issue of Limited Access to Data and Scholarship
Bruce, S. (2018) · The International Information & Library Review · doi:10.1080/10572317.2018.1491708 · Cited by 6
From tedious to targeted: Optimizing oral cancer research with Consensus AI
Pawar, A. M., Desai, R., & Thakur, B. (2024) · Oral Oncology Reports · doi:10.1016/j.oor.2024.100383 · Cited by 4
Research works printed but not linked: 2
- [25]25. Pun, Raymond. 2023. "Generative AI Tools Transforming the Library? Rethinking Possibilities and Questions." Against the Grain 35 (3): 19-22. https://www.charleston-hub.com/2023 /06/generative-ai-tools-transforming-the-library-rethinking-possibilities-and-questions/.
- [32]32. Whitfield, Sharon, and Melissa A. Hofmann. 2023. "Elicit: AI Literature Review Research As- sistant." Public Services Quarterly 19 (3): 201-7. https://doi.org/10.1080/15228959.2023 .2224125. · doi:10.1080/15228959.2023
Cited by
Mohammadkhani, F., & Danesh, F. (2026) · Research Integrity and Peer Review · doi:10.1186/s41073-026-00239-6
Libraries as necessary sites for refusing faux fables of technofascism and imagining otherwise
Appedu, S. (2026) · AI and Ethics · doi:10.1007/s43681-026-01314-w
Using unconventional physical spaces throughout the library for classroom instruction
Miller, K., Korp, A., & George, J. (2026) · The Journal of Academic Librarianship · doi:10.1016/j.acalib.2026.103357
Metadata
- Authors
- Mazel, N.
- Corresponding author
- Mazel, Norah
- Venue
- Library Trends
- Publisher
- Project MUSE
- Published
- 2025-02
- Volume
- 73
- Issue
- 3
- Pages
- 323-343
- ISSN
- 0024-2594 (ISSN-L)
- 1559-0682
- OpenAlex
- W4410829854
- Semantic Scholar
- c89a395811e4f1f4bc6fbbb09634b932cf248376
- Cited by
- 2
- References
- 26
- Citations by year
- 2026: 2
- FWCI
- 6.02
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by-nc-nd
- www.crossref.org/license (vor, 116 d)
- Locations
- Library trends (publishedVersion, cc-by-nc-nd)
- Topics
- Social Sciences › Library and Information Sciences › Library Science and Information Literacy
- Social Sciences › Library and Information Sciences › Library Science and Information
- Social Sciences › Library and Information Sciences › Web and Library Services
- Keywords
- generative AI, information privilege, information literacy instruction, AI research assistant, freemium model, paywalls, access barriers, librarian teaching, AI limitations
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.