because their interests were under-rep- resented in the training data? These are uncomfortable questions, and the reviews touch on them only lightly. They deserve more sustained attention than they have so far received.
Artificial intelligence literacy: a term that needs unpacking
If one phrase runs through nearly every review, it is artificial intelligence literacy. Almost every reviewer reaches for it as the answer to the problems they have just identified. Concerned about hallucinations? AI literacy. Concerned about deskilling? AI literacy. Concerned about over-reliance? AI literacy.
Singapore’s ChatBook service.
The difficulty is that no one has yet established what artificial intelligence literacy actually consists of. Is it technical understanding of how a model works? Is it critical evaluation of outputs? Is it stra- tegic prompting? Is it ethical judgement about when to use these tools and when not? The reviews are vague on the com- ponents, and vaguer still on how artificial intelligence literacy should be taught.
Artificial intelligence literacy has the makings of a useful term, but at present it is doing more rhetorical than analytical work.
StoryGen, horror edition, immersive multimedia experience for library users.
capacity for autonomous action, call for new theory. Until that question is settled, we are working with imperfect maps. The second gap is temporal. Almost all the studies in the reviewed literature are snapshots: surveys taken at a single point in time, or short-term experiments. There are very few longitudinal studies of how an individual’s use of artificial intelligence changes over months or years. Given the speed at which the technology is changing, this is a real problem. We do not know whether the worries patrons express today will fade as the technology improves, or harden as familiarity exposes more of its failure modes.
The third gap is cultural and linguistic. Almost everything in the reviewed liter- ature is in English, conducted in Europe,
18 INFORMATION PROFESSIONAL
North America, and a few other Eng- lish-speaking research locations. When we ourselves searched for reviews in Span- ish, French, German, Arabic, Russian, Hindi, Malay, Indonesian, Japanese, and Chinese, we found one (in Arabic). This is a striking imbalance. The way information behaviour interacts with artificial intelli- gence may look quite different in regions with different infrastructures, languages, and cultural norms about authority and privacy. Most of what is written treats artificial intelligence as if its consequences were universal. They are not. The fourth gap concerns power. Who benefits from these efficiencies? Whose information needs are met by the cur- rent systems, and whose are neglected? Which groups receive worse search results
This matters for libraries, because we are likely to be asked, sooner rather than later, to deliver such programmes; some libraries already are. The risk of running ahead of the research is that we end up teaching whatever the trainer happens to know, which may or may not be what users need. A more measured approach is to begin with the components we can be confident about, namely verification, source checking, recognising fabrication, and prompting effectively, and to add the more contested elements as the research clarifies them. The reviews are honest enough to admit that this clarification has not yet taken place.
What practitioners can do now It would be unsatisfying to end a review of reviews with the message that the research is not yet settled. The literature, taken together, supports five practical points for library practice now.
l Treat artificial intelligence literacy and information literacy as overlapping but distinct. Verification, source checking, and bias awareness apply to both. Prompting and recognising fabrication do not. Build them into existing literacy programmes, rather than running parallel ones;
l be cautious in deploying chatbots. Where they are deployed, prioritise accessibility, plan for emotional-handling failures, and ensure that there is a clear
August-September 2026
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
