CoChat

Navigating the Ethical Challenges of Artificial Intelligence in Higher Education: An Analysis of Seven Global AI Ethics Policies

Open access
Slimi, Zouhaier, Villarejo Carballido, Beatriz · 2023 · TEM Journal

The DOI matches its Crossref record; 18 of the 18 references it cites were found in a registry or an index; none of the 40 sources its claims cite is checked for support yet.

Abstract

AI use in higher education raises ethical concerns that must be addressed. Biased algorithms pose a significant threat, especially if used in admission or grading processes, as they could have devastating effects on students. Another issue is the displacement of human educators by AI systems, and there are concerns about transparency and accountability as AI becomes more integrated into decision-making processes. This paper examined three AI objectives related higher education: biased algorithms, AI and decision-making, and human displacement. Discourse analysis of seven AI ethics policies was conducted, including those from UNESCO, China, the European Commission, Google, MIT, Sanford HAI, and Carnegie Mellon. The findings indicate that stakeholders must work together to address these challenges and ensure responsible AI deployment in higher education while maximizing its benefits. Fair use and protecting individuals, especially those with vulnerable characteristics, are crucial. Gender bias must be avoided in algorithm development, learning data sets, and AI decision-making. Data collection, labeling, and algorithm documentation must be of the highest quality to ensure traceability and openness. Universities must study the ethical, social, and policy implications of AI to ensure responsible development and deployment. The AI ethics policies stress responsible AI development and deployment, with a focus on transparency and accountability. Making AI systems more transparent and answerable may reduce the adverse effects of displacement. In conclusion, AI must be considered ethically in higher education, and stakeholders must ensure that AI is used responsibly, fairly, and in a way that maximizes its benefits while minimizing its risks.

Overview

CoChat Summary

TL;DR

This paper uses discourse analysis of seven global AI ethics policies to examine three higher-education AI ethics themes: biased algorithms, AI decision-making, and human displacement. It concludes that policies converge on fairness, transparency, accountability, and explainability, but implementation and enforcement remain uneven and require stakeholder collaboration.

Key findings

  • Across all seven policies studied—UNESCO, China, the European Commission, Google, MIT, Stanford HAI, and Carnegie Mellon—the paper finds emphasis on guiding AI toward unbiased algorithms and avoiding inequitable effects based on sensitive characteristics such as race, ethnicity, gender, nationality, income, sexual orientation, ability, and political or religious belief.
  • For decision-making, the policies state that gender bias should be avoided or minimised in algorithm development, learning data sets, and AI decision-making; automated decisions using personal information must be transparent, fair, impartial, and non-discriminatory, with explanations and opt-out or refusal rights.
  • For human displacement, the policies argue that when AI decisions affect human life they should be explainable; human interaction with AI should be optional and not impersonated. The European Commission policy calls for responsible deployment, retraining, and social safety nets, while MIT and Stanford policies encourage avoiding displacement and creating new opportunities for workers.
  • The paper’s review of cited empirical work reports, for example, that Yoder-Himes et al. assessed 357 students and found darker-skinned and Black students, and women with the darkest skin tones, were more likely to be flagged for review by automated proctoring software; Fossen et al. found software and industrial robots were related to wage decreases, suggesting displacement effects.

Methods

  • Discourse analysis of seven central AI ethics policies selected through purposive sampling: UNESCO, China, the European Commission, Google, MIT, Stanford HAI, and Carnegie Mellon.
  • The data were thematically coded according to three main themes—AI bias, decision-making, and human labour displacement—and analysed using Fairclough dimensions based on the texts studied.
  • Ethical considerations included ensuring data accuracy and transparency, respecting intellectual property rights, maintaining objectivity and fairness, avoiding negative consequences of recommendations, and adhering to relevant ethical guidelines and regulations.

Limitations

  • Stated: purposive sampling does not accurately reflect the larger population and may limit the generalisability of the findings.
  • Stated: the policies are “still a work in progress and subject to change,” and the level of implementation and enforcement may vary across universities or colleges.
  • Evident from the text: the qualitative discourse analysis of seven policy documents cannot measure actual policy effects or implementation, and the findings rely on selected texts and secondary literature.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.18421/tem122-02

Cites

18 of 18 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Fairclough discourse analysis model
Figure 1 Fairclough discourse analysis modelCC-BY-NC-ND · From the PDF

Verification

Claims the paper makes

0 Valid · 8 Not valid · 35 Not checked · 32 Own results and interpretations

11 more claims

Create a free account to see all 75 claims and to verify the 40 references still open.

See all claims — free account
  • It may state two separate claims as one. 17
  • Probably not a substantial claim. 55

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

5 of 5

Citations

Cites

Show all 18 references

Research works printed but not linked: 1

  1. Fairclough, N. (2009). Critical discourse analysis is the critical study of language in London Longman. Cambridge.

Cited by

See all 314

Metadata

Authors
Slimi, Z., & Villarejo Carballido, B.
Affiliations
  • Universidad de Deusto (ES)
  • Universitat Autonoma de Barcelona, Cerdanyola del Vallès, near the city of Barcelona in Catalonia, Barcelona, Spain
Venue
TEM Journal
Published
2023-05-29
Pages
590-602
ISSN
  • 2217-8309 (ISSN-L)
  • 2217-8333
OpenAlex
W4381570295
Cited by
212
References
18
Citations by year
2026: 70 · 2025: 90 · 2024: 47 · 2023: 4
FWCI
52.80
Type
journal article
Language
en
Open access
Locations
TEM Journal (publishedVersion, cc-by-nc-nd)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
  • Social Sciences › Education › Artificial Intelligence in Education
  • Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
Keywords
AI ethics, higher education, AI ethics policy, algorithmic bias, algorithmic fairness, automated grading, university admission, transparency and accountability, responsible AI deployment, gender bias in AI, vulnerable students

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Navigating the Ethical Challenges of Artificial Intelligence in Higher Education: An Analysis of Seven Global AI Ethics Policies. https://app.cochat.ai/papers/10.18421/tem122-02
The paper
Slimi, Z., & Villarejo Carballido, B. (2023). Navigating the Ethical Challenges of Artificial Intelligence in Higher Education: An Analysis of Seven Global AI Ethics Policies. TEM Journal. https://doi.org/10.18421/tem122-02

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.