AI evidence extraction

Making Artificial Intelligence (AI)–Mediated Translation Visible: A Minimum Disclosure Standard for Qualitative Health Research

Animesh Ghimire · Qualitative Health Research · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1177/10497323261458992

Methodology & findings

Study design

Methodological provocation and conceptual analysis using epistemic injustice and cross-language qualitative scholarship frameworks; includes a vignette analysis demonstrating how AI-mediated translation flattens culturally dense language; proposes a guideline framework (MDS-AIMT) for minimum disclosure standards.

Main result

The paper argues that "the central problem is epistemic: translation technologies are quietly becoming part of the infrastructure of qualitative knowledge production, yet remain methodologically under-disclosed and analytically under-theorised." The author demonstrates "how AI translation can flatten culturally saturated narratives" and shows that without robust human verification, the interpretive commitments of Reflexive Thematic Analysis become difficult to sustain.

Reports effect sizes.

Research paradigm

Interpretivist/Critical realist epistemology grounded in epistemic injustice and reflexive qualitative inquiry

Author conclusions

"My aim is to move the field toward an ethic of epistemic accountability: technology as an assistant, not an author, and translation as an interpretive practice rather than a methodological footnote." The author proposes that "Reflexive Thematic Analysis (RTA) functions as a methodological stress test" for exposing gaps when AI-mediated translation is not robustly verified.

Risk of bias

Not applicable - this is not an empirical study with participants or measurements. As a methodological argument paper, potential bias includes: author's own theoretical commitments to epistemic injustice framework; reliance on single illustrative vignette rather than systematic evidence; lack of empirical validation of the proposed MDS-AIMT guideline.

Open questions raised

The paper identifies that AI translation technologies are becoming normalized in qualitative health research (cross-language interviewing, transcription, and analysis) but remain methodologically under-disclosed and analytically under-theorised. The author calls for greater epistemic accountability in the use of AI translation tools and proposes the MDS-AIMT guideline to address gaps in current practice and disclosure standards.
The paper identifies that AI-powered translation in qualitative health research is "methodologically under-disclosed and analytically under-theorised" despite becoming increasingly normalized. It highlights the need for robust disclosure standards and theoretical frameworks for understanding AI's role in knowledge production.
The paper identifies that AI-powered translation is becoming normalized in qualitative health research (cross-language interviewing, transcription, analysis) yet remains "methodologically under-disclosed and analytically under-theorised." Authors call for development of guidelines and reflexive practices that acknowledge AI as an interpretive agent rather than a neutral tool.

Data: not_statedCode: not_statedExtracted from: pdfAgreement 76%

Explore related topics

Responsible AI Use Academic Writing Position Paper