AI evidence extraction

INTEGRIDADE ACADÉMICA NA ERA DA INTELIGÊNCIA ARTIFICIAL: POLÍTICAS E ESTRATÉGIAS PARA UNIVERSIDADES AFRICANAS

Leonilde Magda Guila Samo · Revista Foco · 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/10.54751/revistafoco.v19n6-085

Methodology & findings

Study design

Qualitative exploratory-descriptive approach using systematic literature review (RSL) combined with documentary analysis of institutional policies and international normative documents.

Sample

> 1000, 3 groups

Primary method

Thematic content analysis (Bardin, 2011) with identification of both deductive and emergent categories. Data processing and organization supported by ATLAS.ti v.23 qualitative analysis software. The systematic literature review followed PRISMA protocol adapted for non-clinical reviews.

Main result

The study reveals that "a maioria das universidades africanas se encontra numa posição de relativa despreparação face a estes desafios, caracterizada pela ausência ou inadequação de políticas específicas, pela insuficiente formação de docentes e estudantes em literacia em IA" (most African universities find themselves in a position of relative unpreparedness to face these challenges, characterized by the absence or inadequacy of specific policies, insufficient training of teachers and students in AI literacy). The analysis of 11 African universities showed that only 2 out of 11 institutions explicitly address artificial intelligence in their academic integrity policies, while others rely on outdated frameworks that do not account for AI-generated text.

Reports effect sizes.

Research paradigm

Critical realism with qualitative interpretative approach

Author conclusions

"A questão que se coloca às universidades africanas não é se devem ou não permitir o uso da IA, mas como podem garantir que a integração desta tecnologia contribua para o desenvolvimento de competências genuínas, para a produção de conhecimento contextualmente relevante e para a formação de cidadãos e profissionais éticos. Responder a esta questão é, em última análise, um acto de afirmação da soberania epistémica africana num mundo em que as agendas tecnológicas são, em grande medida, definidas fora do continente." (The question facing African universities is not whether or not to allow the use of AI, but how they can ensure that the integration of this technology contributes to the development of genuine competencies, to the production of contextually relevant knowledge, and to the training of ethical citizens and professionals. Answering this question is, ultimately, an act of assertion of African epistemic sovereignty in a world where technological agendas are, to a large extent, defined outside the continent.)

Risk of bias

Publication bias: greater representativeness of Anglophone contexts in indexed literature; Language bias: exclusion of non-English, French, and Portuguese documents may omit relevant African scholarship; Digital divide bias: reliance on publicly available online documents may miss policy documents from institutions with limited digitalization; Temporal bias: rapid obsolescence of information in a field undergoing accelerated technological change; Selection bias: analysis limited to 11 African universities may not be representative of the continent's 1,200+ institutions; Publication bias favoring anglophone contexts in indexed literature; Rapid obsolescence of information in a fast-evolving field; Limited access to institutional policy documents from African universities not digitalized or available only in local languages; Potential underrepresentation of non-English language publications; Limited accessibility of African institutional policies not digitized or available only in local languages; Rapid obsolescence of information in a fast-evolving technological field; Potential underrepresentation of African-authored scholarship in major international databases

Limitations

"As limitações desta metodologia incluem a possibilidade de viés de publicação (maior representatividade de contextos anglófonos na literatura indexada), a rápida obsolescência de algumas informações num campo em acelerada transformação e as dificuldades de acesso a documentos de política institucional de universidades africanas não digitalizados ou disponíveis apenas em línguas locais." (The limitations of this methodology include the possibility of publication bias (greater representativeness of anglophone contexts in indexed literature), rapid obsolescence of some information in a field undergoing rapid transformation, and difficulties accessing institutional policy documents from African universities that are not digitized or only available in local languages.)

Open questions raised

The authors recommend future research directions: (a) in-depth case studies in African universities with diverse characteristics to understand factors that facilitate or hinder implementation of academic integrity policies in the AI era; (b) participatory research with African students and faculty on their perceptions, practices and needs regarding AI in academic production; (c) development and validation of learning assessment instruments resistant to AI use and appropriate for specific African contexts; and (d) comparative studies between African universities with different models of academic integrity governance to identify transferable best practices.
The authors recommend future research should include: (a) in-depth case studies in African universities with diverse characteristics to understand facilitating and hindering factors for implementing academic integrity policies in the AI era; (b) participatory research with African students and faculty on their perceptions, practices and needs regarding AI in academic production; (c) development and validation of learning assessment instruments resistant to AI use and suited to specific African contexts; and (d) comparative studies between African universities with different academic integrity governance models to identify transferable best practices.
The authors identify four future research directions: (a) in-depth case studies in African universities with diverse characteristics to understand facilitating and hindering factors in implementing academic integrity policies in the AI era; (b) participatory research with African students and faculty on their perceptions, practices, and needs regarding AI in academic production; (c) development and validation of learning assessment instruments resistant to AI use and appropriate to specific African contexts; and (d) comparative studies between African universities with different academic integrity governance models to identify transferable best practices.

Extracted from: pdfAgreement 59%

Explore related topics

Institutional Policy AI Governance Literature Review