AI evidence extraction

Responsible Artificial Intelligence and Journal Publishing

Shirley Gregor · Journal of the Association for Information Systems · 2024

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

9.51

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.17705/1jais.00863

Methodology & findings

Study design

Opinion piece employing framework analysis and normative governance analysis.

Primary method

Framework design based on synthesis of existing principles and governance structures; normative analysis using design science perspective

Main result

The paper identifies "a high degree of convergence on eight themes for normative principles for responsible AI" including professional responsibility, safety and security, fairness and nondiscrimination, privacy, transparency and explainability, human control of technology, accountability, and promotion of human values. The AI Principled Governance Matrix (AI-PGM) shows "how the whole ecosystem of publishing should be considered when looking at the responsible use of AI—not just journal policy itself, but also coordinating knowledge and action across authors, research organizations, and government legislation."

Research paradigm

Normative/prescriptive (design science and governance frameworks)

Author conclusions

The author concludes: "The new AI-PGM provides a structured means for examining governance practices in terms of the principles and associated risks in the development and use of AI. The matrix shows how the whole ecosystem of publishing should be considered when looking at the responsible use of AI—not just journal policy itself, but also coordinating knowledge and action across authors, research organizations, and government legislation. The AI-PGM has been used here in the context of journal publishing. However, it has the potential to be applied in other fields. As the matrix is new, there are opportunities for others to provide commentary and develop it further."

Risk of bias

Subjective judgment in positioning risks and countermeasures across governance levels; Selective coverage due to rapidly changing AI landscape; Author's personal ethical perspectives may influence framing; Gray literature emphasis (news, reports) which may contain publication bias; Author judgment in categorizing risks and countermeasures across governance levels; Selective coverage of literature due to rapidly evolving field; Potential obsolescence of examples and recommendations

Limitations

The author explicitly acknowledges that "This opinion piece is being written at a time of very rapid change and it is difficult to give comprehensive coverage of the phenomena of interest
Some of what is written is likely to be outdated in the near future." Additionally, the author notes that "the selection of countermeasures is indicative rather than complete, given the rapid developments in the field of AI, and personal judgment has been exercised in placing countermeasures at particular governance levels and against particular principles."

Open questions raised

The author identifies that "more work needs to be done by research institutions to ensure that their codes of academic conduct are being updated" regarding AI use. The paper also notes that "The level of government regulation is experiencing very rapid change, and it is difficult to see what will eventually come to pass." The author invites further development: "As the matrix is new, there are opportunities for others to provide commentary and develop it further."
Need for empirical validation of the AI Principled Governance Matrix (AI-PGM)
Application of the matrix to other fields beyond journal publishing
Further development and refinement of the framework by other researchers
Updates to organizational codes of academic conduct at research institutions
Development of more sophisticated AI detection tools for academic publishing

Extracted from: pdfAgreement 73%

Explore related topics

Responsible AI Use AI Governance Position Paper