CFP: (In)visible AI: What We See, What We Do Not, and Why It Matters

While Artificial Intelligence, from the knowledge graphs that organize the Web to chatbots, is routinely characterized in terms of either its technical capabilities or the dangers it poses, much of what shapes it remains out of view, and thus goes unexamined. Some of the values underpinning these systems—e.g., efficiency, accuracy, convenience—are openly articulated by their producers, while the decisions that shape them—e.g., data selection, labelling guidelines, training, and fundamental operational logic—are kept hidden. Similarly, the labor performed by annotators, moderators, cataloguers, on one hand; as well as that of the writers, artists, and social media users whose contributions become training data, on the other hand, is only seldom acknowledged.

Yet visible and invisible aspects alike factor into, operationally—and thus determine in practice—whose values and perspectives become embedded in AI, which ones it enacts, and therefore what counts within those systems—what gets measured, and what matters. This Special Issue leverages the invisibility-visibility dichotomy and treats it as the common thread that unites the many facets of AI systems that are themselves both visible—e.g., a chatbot—and invisible—e.g., a customer care virtual assistant disguised as a human—depending on the context.

We invite contributions grounded in a single, concrete case—e.g., a dataset, an interface, a mechanism, a work/research context or site—that bring the invisible into view.

Potential questions

• • • Whose work supports AI operations? Which kinds of work are visible, and which are invisible? Why?
    • What metrics are considered when AI systems are evaluated? Whose values underpin these metrics?
    • What kinds of materials, sources, ways of knowing are out of reach of AI?
    • How do we know we are interacting with an AI system, and when are we not told?
    • How do knowledge graphs and ontologies shape what information AI can retrieve and leverage?
    • How do transparency reports foreground some information while keeping other aspects from view?
    • How does the language we use to describe AI shape what we see in, and expect from, it? What does this language foreclose?
    • When can invisibility from AI systems be a form of protection or refusal?

Author Instructions

All proposals should be submitted directly through the Information Matters’ platform, following the author’s instructions. Authors are encouraged to include illustrations with their texts. Submitted pieces should ideally be 1,000 words, aimed at a general audience, and should not contain traditional academic citations or reference lists. Instead, authors may embed links where appropriate.

Before submitting, please create an account at https://informationmatters.org/. Once registered, create a post that will serve as your submission to the Special Issue. When submitting your article, make sure to designate “SI Invisible AI” as a tag.

For your writing reference, please find published articles here in the recently published Special Issue, Information Literacy: A Discipline for the Future.

Timeline

Submission deadline: December 15, 2026
Full issue publication: January 2027

Decisions and publication via the Information Matters website: typically within two weeks of submission.

Questions
Please send your questions about this Special Issue to Pietro Santachiara: psantachiara@ucla.edu

Guest Editor
Pietro Santachiara, MLIS
PhD Candidate in Information Studies
University of California, Los Angeles