nucleation

LIT-1338: Epistemic Injustice in Generative AI

1 min read

Properties5
StatusActive — read 2026-10-21 (NOTE-1049) in full, from the publisher's free PDF of the AIES 2024 proceedings; worth reading for its four-way taxonomy of how generative models take part in epistemic injustice, and for the point that amplified testimonial injustice runs through a credibility excess assigned to the model (THEORY-864). Read it knowing that it is conceptual and measures nothing, that its examples are others' reports, that its "manipulative" configuration counts deliberate fabrication as testimonial injustice although Fricker sets deliberate manipulation apart, and that its access injustice stretches the hermeneutical category.
Version2
Tagsepistemology, ethics, anthology-candidate
Cited as “paper” byNOTE-1049 — Kay, Kasirzadeh and Mohamed — Epistemic injustice in generative AI
Cited as “source” byTHEORY-864 — When a generative model reproduces a prejudiced narrative learned from its training data, the resulting testimonial injustice runs through a credibility excess assigned to the model, so that hearers believe the amplified account over the testimony of those it discredits, rather than through a hearer's prejudiced judgement of a present speaker

Graph View

Kay, J., Kasirzadeh, A. & Mohamed, S. (2024). Epistemic Injustice in Generative AI. Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society 7: 684–697. doi:10.1609/aies.v7i1.31671. Preprint: arXiv:2408.11441 (21 August 2024).

Read in NOTE-1049, which records what the paper argues and where it departs from Fricker.

Why this is here

A25 tied it to the record itself: “What becomes searchable, summarized, quoted, attributed, or incorporated into future reasoning may influence what subsequently becomes authoritative.” Bears on CLAIM-406, beside the record’s Fricker readings (THEORY-317, THEORY-327).

Backlinks