Long-Term Preservation Challenges of Large Language Models and Their Ephemeral Outputs in Archives in South West Nigeria

References

[1] Albani, M., & Maggio, I. (2020). Long-term data preservation: Data lifecycle, standardisation process, implementation and lessons learned. International Journal of Digital Curation, 15(1), Article 715. https://doi.org/10.2218/ijdc.v15i1.715

[2] Areo, P. O., Aliyu, A., & Jega, A. M. (2023). Perspectives in preservation of digital materials in libraries: Skills and competencies. International Journal of Research and Innovation in Social Science, 7(69), 1096–1101. https://doi.org/10.47772/IJRISS.2023.7693

[3] Brichkovsky, V. (2023). Challenges and prospects of long-term preservation of digital cultural heritage. Bulletin of Armenian Libraries, 6(1), 19–22. https://doi.org/10.52027/journal.nla.am

[4] Carlini, N., Tramèr, F., Wallace, E., Jagielski, M., Herbert-Voss, A., Lee, K., Roberts, A., Brown, T., Song, D., Erlingsson, Ú., Oprea, A., & Raffel, C. (2021). Extracting training data from large language models. In Proceedings of the 30th USENIX Security Symposium (pp. 2633–2650). USENIX Association. https://www.usenix.org/conference/usenixsecurity21/presentation/carlini-extracting

[5] Consultative Committee for Space Data Systems (CCSDS). (2012). Reference model for an open archival information system (OAIS): Recommended practice (CCSDS 650.0-M-2). CCSDS Secretariat. https://public.ccsds.org/Pubs/650x0m2.pdf

[6] Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). SAGE Publications.

[7] Di Cosmo, R., & Zacchiroli, S. (2017). Software heritage: Why and how to preserve software source code. In iPRES 2017: Proceedings of the 14th International Conference on Digital Preservation (pp. 1–10). https://hal.science/hal-01590958

[8] Gebru, T., Morgenstern, J., Vecchione, B., Vaughan, J. W., Wallach, H., Daumé III, H., & Crawford, K. (2021). Datasheets for datasets. Communications of the ACM, 64(12), 86–92. https://doi.org/10.1145/3458723

[9] Khalid, A. (2022). Conservation challenges and emerging trends of digital preservation for UNESCO architectural heritage, Pakistan. Conservation, 2(1), 26–37. https://doi.org/10.3390/conservation2010003

[10] Kitchin, R. (2014). The data revolution: Big data, open data, data infrastructures and their consequences. SAGE Publications.

[11] Lavoie, B. F. (2014). The open archival information system (OAIS) reference model: Introductory guide (2nd ed., DPC Technology Watch Report 14-02). Digital Preservation Coalition. https://doi.org/10.7207/twr14-02

[12] Masenya, T. M., & Ngulube, P. (2019). Digital preservation practices in academic libraries in South Africa in the wake of the digital revolution. South African Journal of Information Management, 21(1), a1011. https://doi.org/10.4102/sajim.v21i1.1011

[13] Mitchell, M., Wu, S., Zaldivar, A., Barnes, P., Vasserman, L., Hutchinson, B., Spitzer, E., Raji, I. D., & Gebru, T. (2019). Model cards for model reporting. In Proceedings of the Conference on Fairness, Accountability, and Transparency (pp. 220–229). ACM. https://doi.org/10.1145/3287560.3287596

[14] Paullada, A., Raji, I. D., Bender, E. M., Denton, E., & Hanna, A. (2021). Data and its (dis)contents: A survey of dataset development and use in machine learning research. Patterns, 2(11), 100336. https://doi.org/10.1016/j.patter.2021.100336

[15] Pineau, J., Vincent-Lamarre, P., Sinha, K., Larivière, V., Beygelzimer, A., d’Alché-Buc, F., Fox, E., & Larochelle, H. (2021). Improving reproducibility in machine learning research (A report from the NeurIPS 2019 reproducibility program). Journal of Machine Learning Research, 22(164), 1–20. http://jmlr.org/papers/v22/20-303.html

[16] Raji, I. D., Smart, A., White, R. N., Mitchell, M., Gebru, T., Hutchinson, B., Smith-Loud, J., Theron, D., & Barnes, P. (2020). Closing the AI accountability gap: Defining an end-to-end framework for internal algorithmic auditing. In Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency (pp. 33–44). ACM. https://doi.org/10.1145/3351095.3372873

[17] Ubogu, J. O., Ejimonye, O. C., Hubert-Nwangwu, O., Ekeh, U. C., & Ijantiku, C. M. (2023). Digital preservation challenges in the era of big data: Strategies for long-term information storage in libraries. International Journal of Academic Library & Information Science, 11(7), 249–253. https://doi.org/10.14662/ijalis2023270

[18] United Nations Educational, Scientific and Cultural Organization (UNESCO). (2021). Recommendation on the ethics of artificial intelligence. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000381137

@article{1718818,
      author = {Dada, Rachael Olufunmilola, Adebowale Jeremy Adetayo, Akintunde, Ebunoluwa Boluwatife, Peter Enamudu},
      title = {Long-Term Preservation Challenges of Large Language Models and Their Ephemeral Outputs in Archives in South West Nigeria},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {9},
      number = {12},
      pages = {4072-4081},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1718818.pdf},
      abstract = {Archival institutions worldwide are confronting transformative pressures as Large Language Models (LLMs) generate content that challenges established preservation frameworks. This study examined the long-term preservation challenges of LLMs and their ephemeral outputs in archival institutions in South West Nigeria. A descriptive survey design was adopted, and structured questionnaires were administered to 71 archival and information management professionals drawn from National Archives, state archives, academic institutions, and government agencies across Ogun, Oyo, Lagos, and Ekiti States. Findings reveal that while awareness of LLM technologies is high (93%), institutional adoption remains uneven, with 33.8% of institutions never using LLMs. Respondents largely characterise LLM outputs as dynamic, context-dependent, and continuously updated artefacts. A majority (59.2%) conditionally recognise these outputs as official records, with recognition contingent on usage context. Key preservation challenges identified include unstable power supply and poor internet connectivity (M = 3.17), lack of technical expertise (M = 3.06), inadequate funding (M = 3.01), and the absence of AI-specific preservation policies (M = 3.01). Although most institutions possess general digital preservation policies, guidelines specifically addressing AI-generated content remain critically underdeveloped, and over half of professionals (56.3%) have not received formal AI preservation training. The study recommends adoption of international preservation standards, targeted capacity building, sustained funding, and integration of AI-generated records into national archival frameworks. These findings advance understanding of digital preservation challenges in emerging archival ecosystems and have direct relevance to archival policy in sub-Saharan Africa.},
      keywords = {archives, artificial intelligence, digital preservation, ephemeral outputs, large language models},
      month = {June},
      doi = {https://doi.org/10.64388/ire1718818}
  }