possible. There are a number of reasons why open access repositories and library catalogues are so attractive and there- fore heavily targeted. First, they are open by their very nature (not behind paywalls or authentication systems), the content is peer-reviewed (written by experts), and the metadata is structured (and therefore easy to interpret). Problematically for us main- tainers, this data has rich links (subject, author headings, etc.) which the AI tools can pretty much continue indefinitely fol- lowing ultimately bringing servers to a halt. Historically, and to this day, our library catalogues and repositories maintain a robots.txt file (
https://en.wikipedia.org/wiki/ Robots.txt) which essentially says we are happy (or not) to give permission for our collections being indexed. The AI crawlers we are seeing are ignoring this file entirely and retrieving the content with or without permission. I have also seen similar articles about how the creative industry is con- cerned about copyright theft. There are still unanswered questions around stolen con- tent by AI tool creators, which will no doubt continue to play out in court cases over the coming years (
https://tinyurl.com/twoAIcases). Another library I spoke to recently asked about ‘pay per crawl’. There are ways of doing this now but, hopefully, as some rules and regulations start to emerge around AI practice, there will also emerge opportuni- ties to leverage content, particularly content which is in open access repositories (or just library catalogues). With library budgets so tightly stretched, why shouldn’t libraries be able to ask AI companies to pay for each crawl of their content?
Digital sovereignty
Digital sovereignty is a topic I think we will hear more and more about in the coming years. Much of the technology we use in our day to day lives comes from big tech which are largely sovereign outside of the UK (primarily in the US). Our world feels like a fragile place right now, where there seems to be more division than cooperation. How comfortable do we feel being tied into this technology and how would we manage if it was turned off tomorrow? I recently wrote an article about this topic (
https://openfifth.co.uk/whats- on-our-mind-digital-sovereignty-open-source-pub- lic-services/) and highlighted the important role that open data and open source plays in achieving digital sovereignty. In the EU, there’s a growing movement towards ‘digital sovereignty’ or ‘digital autonomy’. This was specifically driven by a desire to reduce reliance on non-European tech giants for critical digital infrastructure and services. The European Commission’s ‘Open source software strategy’ 2020- 2023 (
https://tinyurl.com/EUopensourcestrategy) said ‘Think Open’. I use my local farm shop as I am supporting my local econ- omy, I know it will be there if there were ever to be food supply issues and I pretty much know where everything has come from. Will the same become true of our
Anubis web traffic between May and October 2025 from a Koha system. 2.67 per cent in green represents users / bots who successfully proved they were legiti- mate – everything else is suspicious behaviour.
digital consumption? If these tech giants ever went away, would we still have a sus- tainable, local technology infrastructure? The recent Cloudflare outage also high- lighted our reliance on these US companies. Although the outage was relatively short (around two hours for us), Cloudflare now protects around 20 per cent of internet traffic, including library applications like our own. We actually use both Cloudflare and Anubis (
https://github.com/TecharoHQ/an- ubis) to protect our sites from an influx of largely AI traffic, which meant the impact was a bit more distributed, but it does highlight the double-edged sword of one group of tech companies firstly bringing down systems by bombarding them, and then the company brought in to protect against this also bringing systems to a standstill!
AI in library applications So, where does this leave us regarding using AI inside of library systems? I guess the very obvious one is Natural Language Searching (NLS) which is quickly emerging in library discovery platforms like EBSCO EDS and TDNet Discover. This allows users to search for queries like ‘I want a book about Sweden for my upcoming holiday’ rather than searching for a specific title. Of course, we don’t have to use AI to do this. Aspen Discovery, for example, offers NLS but via its own internal algorithms. How- ever, I think this is an obvious use case. Another area is the emergence of AI cataloguing tools. We are starting to see library systems adding the ability to create catalogue records using AI tools. This is abhorrent to some librarians I have spoken to, but it will likely become more widely adopted in the coming years. Going for- ward, there is an opportunity with retro-cat- aloguing (which I’m sure many libraries have plenty of!). Report queries and analytics is also an obvious place to add AI. Koha already has a ‘Reports Assistant’ (https://chatgpt. com/g/g-Uy73aJR6R-koha-report-assistant). The big benefit of this is that the assistant understands the database schema of Koha
and can talk the user through creating a report. We find our library users have quite a varying degree of SQL knowledge, but the Reports Assistant can help with basic queries like ‘What is the SQL to find overdue items?’ to more advanced queries like ‘My report is running slowly, can you help optimise my SQL query?’ In terms of accessibility, AI is already assisting in identifying (and helping to fix) accessibility issues with websites, which could feasibly be extended to library web- sites and online catalogues. This is a use case where there are already strictly laid down rules for an AI tool to follow. Linked to this, Koha is already translated into over 40 languages and all these require mainte- nance when new versions emerge. AI tools, despite not being quite nuanced enough yet, could assist doing base level translations supported by manual (human) checking. Finally, collection management. It’s possi- ble that AI can assist in making data-driven decisions based on usage about underused collections or gaps within collections.
Summary
As a technology company we are, and have to be, looking at AI. However, we don’t want to produce AI tools for the sake of AI tools. We really want to produce AI tools that add actual value and that have also been ethically and environmentally thought through. There are clearly societal benefits of introducing AI, particularly to speed up and automate organisations that are drowning in admin and paperwork. However, the role of libraries and librar- ians will remain hugely important. There are many in society without information literacy skills and awareness, people may not know that the responses they are receiving from AI tools aren’t good, lack quality, contain hallucinations, etc. We are excited about AI but still believe that our libraries are here to help people nav- igate their way through this new world. They represent everything that is good about information! BG
l
https://openfifth.co.uk/ 9
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62
