AI evidence extraction

AI Disclosure without Accountability: Paper Compliance and the Governance Limits of Transparency in Scientific Research

Victor Frimpong · International Journal of Social Science Studies · 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/10.11114/ijsss.v14i3.8747

Methodology & findings

Study design

Exploratory review of 80 recent academic articles examining patterns of AI disclosure in scientific research and analyzing the gap between reported and actual AI use in research outputs.

Sample

N = 80, 1 group

Primary method

Qualitative pattern analysis; no quantitative statistical testing reported. Descriptive review methodology without formal hypothesis testing.

Main result

The study identifies patterns suggesting that "explicit AI disclosure remains limited and, when present, is primarily symbolic or narrative rather than verifiable." The paper introduces the concept of the "AI Disclosure Integrity Gap (AIDG), defined as the discrepancy between reported AI use and the actual epistemic influence of AI on research outputs," revealing that institutionalising AI disclosure has led to compliance that prioritises symbolic transparency over genuine accountability.

Reports effect sizes.

Research paradigm

Critical interpretivism

Author conclusions

The authors conclude that "disclosure as a necessary but insufficient mechanism" for AI governance, and highlight "the risk that current practices may create a false sense of transparency." They position the AI Use Traceability Framework as a process-oriented governance model that emphasises traceability, auditability, and transparency to strengthen accountability in AI-assisted research.

Risk of bias

Potential selection bias in article sampling (80 articles); non-standardized assessment criteria across reviewed articles; potential publication bias in which articles disclose AI use.; Selection bias in article sampling (80 articles analyzed from undefined population); Potential confirmation bias in identifying patterns supporting the main thesis; No inter-rater reliability reported for qualitative coding; Limited transparency on article selection criteria

Limitations

The paper acknowledges "practical, technical, and institutional constraints associated with implementing traceability mechanisms" and recognizes that while disclosure is "necessary but insufficient," current practices "may create a false sense of transparency" rather than addressing root causes of accountability gaps.

Open questions raised

The paper identifies the need for stronger accountability mechanisms beyond symbolic disclosure, implementation of process-oriented traceability frameworks, and development of standardised, verifiable AI disclosure protocols in scientific publishing.
The paper identifies gaps between disclosure-oriented governance frameworks and the actual characteristics of AI-assisted knowledge production, particularly regarding the iterative, opaque, and often irreproducible nature of AI tools. It calls for strengthened accountability mechanisms beyond symbolic compliance.
The paper identifies the need for process-oriented governance models that move beyond disclosure-based approaches. The authors highlight gaps in understanding how to implement traceability, auditability, and transparency mechanisms in practice, and call for stronger accountability frameworks that address the AI Disclosure Integrity Gap (AIDG).

Data: not_statedCode: not_statedExtracted from: pdfAgreement 61%

Explore related topics

Responsible AI Use Research Integrity Content Analysis