Showing posts with label Algorithmic Bias. Show all posts

When Machines Judge Their Makers: A Critical Appraisal of Turnitin’s AI Detection in Academic Integrity

Academic Integrity, AI Detection, Algorithmic Bias, Ethics, False Positives, Higher Education Ethics, Turnitin 0 comments

Opaque AI judgment in Academia

AI-generated picture by Prof. Jonathan Acuña Solano in November 2025

Introductory Note to the Reader

After a conversation I had with my friend, Dr. Alberto Delgado, a language professor working for the School of Modern Languages at the University of Costa Rica, I decided to write this entry for my blog.

Alberto told me that a paper of his, which he was intending to publish through a magazine issued by Universidad Nacional Autónoma (UNA), the second most important state university in Costa Rica, had been rejected outright. A representative of the evaluators informed him that his paper had been flagged while using Turnitin, and that it would not be considered for publication.

Alberto was not given a hearing, an explanation, or an oral examination to verify authorship. At portas, he was simply told that his paper had been AI-generated. What kind of appraisal was this for a language professor with over 30 years of teaching experience?

When Machines Judge Their Makers: A Critical Appraisal of Turnitin’s AI Detection in Academic Integrity

Abstract

This essay critically examines Turnitin’s AI writing detection system, focusing on its technical flaws, ethical implications, and procedural misuse in educational and academic publishing contexts. Drawing on personal experience—as well as documented cases of false positives, algorithmic bias, and institutional overreliance—the essay argues that Turnitin’s AI module functions as an opaque and unreliable evaluative mechanism. Its deployment as quasi-judicial evidence in academic integrity procedures risks harming legitimate writers, especially non-native English users and scholars whose work predates modern AI systems. The discussion concludes with recommendations for transparent, human-centered approaches to authorship verification and more responsible integration of AI in academic environments.

Keywords:

Turnitin, AI Detection, Academic Integrity, False Positives, Algorithmic Bias, Higher Education Ethics, Ethics

Resumen

Este ensayo ofrece un análisis crítico del sistema de detección de escritura generada por IA de Turnitin, destacando sus fallas técnicas, implicaciones éticas y uso indebido dentro de procedimientos académicos y editoriales. A partir de experiencias personales y evidencia documentada sobre falsos positivos, sesgos algorítmicos y dependencia institucional excesiva, se argumenta que el módulo de IA de Turnitin opera como un mecanismo evaluativo opaco y poco confiable. Su uso como “prueba” en procesos de integridad académica pone en riesgo a escritores legítimos, especialmente a usuarios de inglés como lengua extranjera y autores cuyas obras anteceden a los sistemas actuales de IA. El texto finaliza con recomendaciones para enfoques más humanos, transparentes y justos en la verificación de autoría dentro de la academia.

Resumo

Este ensaio analisa criticamente o sistema de detecção de escrita gerada por IA do Turnitin, enfatizando suas falhas técnicas, implicações éticas e uso inadequado em contextos acadêmicos e editoriais. Com base em experiências pessoais e em casos documentados de falsos positivos, viés algorítmico e dependência institucional excessiva, argumenta-se que o módulo de IA do Turnitin funciona como um mecanismo avaliativo opaco e pouco confiável. Seu uso como evidência em processos de integridade acadêmica coloca em risco autores legítimos, especialmente aqueles que escrevem inglês como segunda língua ou que produziram textos muito antes do surgimento das IAs modernas. O ensaio conclui com recomendações para abordagens mais humanas, transparentes e responsáveis na verificação de autoria.

Introduction

In recent years, educational institutions have rushed to deploy algorithmic “solutions” to concerns about generative AI tools (e.g., ChatGPT, Claude AI, Gemini, Deep Seek, etc.) being used to produce student work. One prominent example is Turnitin’s AI writing detection module, which claims to identify text likely produced by an AI rather than a human (Turnitin, n.d.-c). While the goal of preserving academic integrity is legitimate, the implementation of such detection as “quasi-judicial evidence” is deeply problematic. Cases have already emerged of students being falsely flagged as AI-generated texts. In my very own experience, Turnitin has flagged essays I composed in 2010, long before the popularization of AI writing systems, as AI-generated, but how come? Such errors cast doubt not only on Turnitin’s technical claims, but on the ethics of using such a system to police human scholarship and penmanship.

The intention of this essay is to examine the major deficiencies of Turnitin’s detection approach: (1) its susceptibility to false positives, (2) opacity and lack of verifiability, (3) algorithmic bias, and (4) misuse in academic procedures. Finally, it outlines recommendations for more humane, transparent, and just approaches to dealing with AI in education.

False Positives: When Human Writings Are Falsely Flagged

One of the gravest defects in Turnitin’s AI detection is its tendency to flag purely human-authored content as AI-generated. Turnitin refers to such misclassifications as false positives, meaning the system labels human text as AI. The company itself acknowledges a small, but nonzero, false positive rate (less than 1 %) under ideal conditions (Turnitin, n.d.-a). So, is my writing style curated over more than 30 years of work at the university level misclassified as false positive?

Independent analyses suggest the false positive rate is much higher and context-dependent. Reports have indicated “higher incidence of false positives” when less than 20% of a document is flagged as AI-generated (K-12 Dive, 2023). Empirical testing of AI detectors, including Turnitin’s, has shown that error rates can exceed 10% in uncontrolled settings (Weber-Wulff et al., 2023). In practice, as in my own situation, decades-old writing, long predating modern AI systems, can trigger the detector. That fact alone undermines any claim that the system reliably differentiates AI versus human provenance. If a tool misfires on known human work, its verdicts on ambiguous texts carry no weight.

Opacity and Lack of Verifiability

Beyond error rates, Turnitin’s AI detection suffers from conceptual opacity. Unlike its plagiarism component, which highlights matched passages and links to original sources, the AI detection report offers no “traceback” to a source of suspicion. The system does not allow instructors or students to verify which phrases or sentences prompted the AI label (Salem, Fiore, Kelly, & Brock, n.d.). Because there is no “original text” to which flagged content can be traced, users are left entirely in the dark about why the system made its call.

That opacity severely weakens any claim that the system used by Turnitin is fair or evidence-based in academic adjudication. A student cannot counter or refute the detection logic, especially when the system itself offers no human-readable rationales. In an adjudicative environment, suspicion without scrutiny is a violation of procedural fairness especially if learners cannot defend their positions regarding Turnitin’s verdict.

Algorithmic Bias and Disparate Impact

AI detection systems, including Turnitin’s, may also exacerbate existing inequalities. Several studies suggest that non-native English writers or writers with simpler or more repetitive style are disproportionately flagged (The Markup, 2023). Controlled tests across multiple detectors found that non-native English writing was misclassified as AI-generated in over 60% of cases, whereas native English texts rarely triggered such misclassification (Stanford Human-Centered Artificial Intelligence [HAI], 2023). In my particular case, I’m not a native speaker but someone with C2, better than many educated native speakers, so my texts are flagged as if I have generated them with AI.

A cross-detector study by Liang, Yuksekgonul, Mao, Wu, & Zou (2023) confirmed systematic bias: non-native English writers’ texts were more likely misclassified as AI than native writers’ texts, even when both were human-authored. This finding suggests that algorithmic decisions are entangled with linguistic privilege, penalizing students already disadvantaged by language or cultural background.

Although Turnitin has responded that its detector shows “no statistically significant bias” for English-language learners in certain internal tests (Turnitin, n.d.-b), the lack of transparency regarding its data sets and methods makes such claims unverifiable.

Misuse in Academic Procedures: The Perils of Reliance

When an imperfect, opaque system is elevated to a quasi-authority in academic integrity processes, its defects become not merely technical flaws but instruments of injustice. Institutions sometimes use Turnitin’s AI flag as “prima facie” evidence of misconduct, shifting the burden onto students to “prove their innocence.” In one striking case, the Australian Catholic University acknowledged that Turnitin’s AI tool led to false accusations against students, delaying graduations and causing distress. The institution eventually ceased relying on the tool when used in isolation (Adelaide Now, 2023). Similarly, Vanderbilt University disabled Turnitin’s AI detector entirely, citing concerns about its error rates and lack of transparency (Vanderbilt University Center for Teaching, 2023).

Relying on Turnitin as if it were “infallible” encourages faculty to outsource judgment rather than engage with student writing, context, drafts, and meaning. As a seasoned educator working as a language teacher for over 35 years, relying on AI to do one’s job is an example as to how certain types of teaching professionals degrade pedagogy and erode trust, turning the classroom into a surveillance zone. And what can be said when a paper is submitted to be published in a magazine and it is “checked” for AI generation with Turnitin? Isn’t it another example of outsourcing judgment?

Ethical and Epistemological Objections

At a more fundamental level, what can be said is that policing writing provenance via black-box algorithms is a way to betray an already misguided epistemology. Writing is not a binary product of “human vs. AI,” especially in an era where humans increasingly rely on digital tools such as dictionaries, grammar checkers, and translation aids. The insistence on policing a metaphysical boundary between “just human” and “AI-assisted” is naive and reductive. AI-generated texts are not being “defended” here, the sole intention of this paper is to make teachers help learners use AI to help them with their work such as a cohesion checker, not as a “term-paper producer,” which is unethical.

Furthermore, the ethical consequences of false accusation can be severe: damage to reputation, academic record, emotional distress, and even expulsion. The risk of harm weighs heavily against delegating moral judgment to a fallible system, something that is simply outrageous. And what about those professors wanting to have an article or research paper published where AI was used to improve mechanics, coherence, word choice, data analysis, and so on? Should they be flagged because ChatGPT, Claude.AI, etc. were used?

Recommendations

From my personal stance, as someone who has been working with English language students predating the advent of AI, these are some of my recommendations to help teaching professionals better cope with AI use in paper or research writing:

Limit Use as Heuristic Only	The AI detection score should serve only as a trigger for inquiry, not as conclusive evidence. Faculty should always combine the tool’s output with qualitative judgment, drafts, revision history, and the submitter’s explanation.
Mandate Transparency	Turnitin must make its detection logic, at least at a high level, public so that flagged authors or educators can meaningfully interrogate the decision and then have a second opinion.
Independent Audit and Validation	Universities should commission independent testing of Turnitin’s AI detector on diverse corpora, including high-variance styles, non-native writers, and older texts, texts produced by faculty members, and so on.
Opt-Out or Appeal Rights	Text writers should have the right to contest AI flagging, have their writing re-evaluated by human committees, and demand evidence beyond a single opaque score.
Pedagogical Redesign	Rather than rely on policing, courses and assessments should evolve to emphasize process, draft-based assignments, oral defenses, and in-class writing, formats less vulnerable to AI misuse.
Phase Out Flawed Detection	In institutions already seeing abuse, Turnitin’s AI detector should be disabled or de-emphasized until it can meet rigorous transparency and fairness standards to stop flagging authors wrongly.

Conclusion

Turnitin’s AI writing detection tool positions itself as a guardian of academic integrity, but in truth it is a blunt, opaque, and potentially prejudicial instrument in the hands of people who simply want to outsource judgment instead of doing their job: “reading the author’s paper.” It is already known that it can be susceptible to false positives, contains what experts have labeled as algorithmic bias, and its misuse in academic adjudication threatens to punish honest non-native writers whose C1 or C2 levels are being questioned. My own experience of being flagged for academic writing composed in 2010, while taking a course at Homerton College, University of Cambridge, starkly illustrates how the system can misfire a false positive disastrously.

Academia must resist the temptation to outsource ethical judgment to algorithms. Until such detection systems become transparent, independently validated, and procedurally constrained, they should serve only as a “flag”, not a “guilty verdict”. The responsibility for assessing writing must rest with human educators in dialogue with writers, not buried in binary scores from inscrutable machines.

📚 References

Adelaide Now. (2023, June 14). ‘Robocheating’ fiasco saw Australian Catholic University students falsely accused of using AI by an unreliable AI tool. News Corp Australia. https://www.adelaidenow.com.au/education/higher-education/robocheating-fiasco-saw-australian-catholic-university-students-falsely-accused-of-using-ai-by-an-unreliable-ai-tool/news-story/4a08732c84499263a709ec3bb1980802

K-12 Dive. (2023, June 7). Turnitin admits there are some cases of higher false positives in AI writing detection tool. https://www.k12dive.com/news/turnitin-false-positives-ai-detector/652221

Liang, W., Yuksekgonul, M., Mao, Y., Wu, E., & Zou, J. (2023). GPT detectors are biased against non-native English writers [Preprint]. arXiv. https://arxiv.org/abs/2304.02819

Salem, L., Fiore, S., Kelly, K., & Brock, B. (n.d.). Evaluating the effectiveness of Turnitin’s AI writing indicator model. Temple University Center for the Advancement of Teaching. https://teaching.temple.edu/sites/teaching/files/media/document/Evaluating%20the%20Effectiveness%20of%20Turnitin%E2%80%99s%20AI%20Writing%20Indicator%20Model.pdf

Stanford Human-Centered Artificial Intelligence (HAI). (2023, May 1). AI detectors biased against non-native English writers. https://hai.stanford.edu/news/ai-detectors-biased-against-non-native-english-writers

The Markup. (2023, August 14). AI detection tools falsely accuse international students of cheating. https://themarkup.org/machine-learning/2023/08/14/ai-detection-tools-falsely-accuse-international-students-of-cheating

Turnitin. (n.d.-a). Understanding false positives within our AI writing detection capabilities. https://www.turnitin.com/blog/understanding-false-positives-within-our-ai-writing-detection-capabilities

Turnitin. (n.d.-b). New research: Turnitin’s AI detector shows no statistically significant bias against English language learners. https://www.turnitin.com/blog/new-research-turnitin-s-ai-detector-shows-no-statistically-significant-bias-against-english-language-learners

Turnitin. (n.d.-c). Does Turnitin detect AI writing? Debunking common myths and misconceptions. https://www.turnitin.com/blog/does-turnitin-detect-ai-writing-debunking-common-myths-and-misconceptions

Vanderbilt University Center for Teaching. (2023, August 16). Guidance on AI detection and why we’re disabling Turnitin’s AI detector. Vanderbilt Brightspace Blog. https://www.vanderbilt.edu/brightspace/2023/08/16/guidance-on-ai-detection-and-why-were-disabling-turnitins-ai-detector

Weber-Wulff, D., Anohina-Naumeca, A., Bjelobaba, S., Foltýnek, T., Guerrero-Dib, J., Popoola, O., & Waddington, L. (2023). Testing of detection tools for AI-generated text [Preprint]. arXiv. https://arxiv.org/abs/2306.15666

Reader's Handout - Interactive Reading & Reflection Guide

Reader's Handout - Interactive Reading & Reflection Guide by Jonathan Acuña

When Machines Judge Their Makers by Jonathan Acuña

Wednesday, November 26, 2025

The Risks of Misusing AI in Costa Rican Education

Algorithmic Bias, Artificial Intelligence, Costa Rica, Digital Inequality, Education Policy, Teacher Training 0 comments

Working with AI

AI-generated picture by Prof. Jonathan Acuña Solano in August 2025

✍️ Introductory Note to the Reader

I am not entirely sure how I came across the article AI Can Revolutionise Education but Technology Is Not Enough: Human Development Meets Cultural Evolution. Yet, as I read through it, I was struck by the comparison between Uruguay’s forward-looking integration of AI in education and the slower pace that Costa Rica is likely to experience—perhaps over decades.

This contrast left me with a sense of sadness, especially as an English teaching professional who witnesses daily how students often misuse AI. Many use it to bypass homework, avoid thinking in the target language, or escape the cognitive effort of practicing English. As I often remind my students: you will not have subtitles or “zapping” in a real job interview. AI can indeed be a powerful educational tool, but the critical question remains—who truly knows how to use it responsibly and effectively?

The Risks of Misusing AI in Costa Rican Education

Abstract

Artificial Intelligence (AI) is reshaping education worldwide, offering personalized learning, automated assessment, and expanded access. In Costa Rica, these opportunities intersect with a strong national commitment to human development and equity. However, the misuse of AI threatens to undermine rather than enhance educational outcomes. This essay examines the risks of uncritical AI integration in Costa Rican universities, language institutes, and public high schools. Key dangers include student overdependence on AI tools, algorithmic bias and cultural mismatch, widening digital inequality, weak teacher preparation, and ethical concerns related to data privacy. Drawing on Muthukrishna et al. (2025), Darvishi et al. (2024), and others, the essay argues that technology-first approaches replicate past failures, while successful strategies must embed AI in systems that prioritize infrastructure, pedagogy, cultural adaptation, and teacher empowerment. Without these safeguards, AI could erode autonomy, equity, and cultural relevance in Costa Rican education.

Keywords:

Artificial Intelligence, Education Policy, Costa Rica, Digital Inequality, Algorithmic Bias, Teacher Training

Resumen

La inteligencia artificial (IA) está transformando la educación global, y Costa Rica no es la excepción. No obstante, el uso inadecuado de estas tecnologías puede generar efectos negativos en lugar de potenciar el aprendizaje. Este ensayo analiza los principales riesgos de la implementación acrítica de la IA en universidades, institutos de idiomas y colegios públicos costarricenses. Se identifican como peligros centrales la dependencia excesiva de los estudiantes en las herramientas de IA, el sesgo algorítmico y la falta de adecuación cultural, la brecha digital entre zonas urbanas y rurales, la escasa preparación docente y las preocupaciones éticas relacionadas con la privacidad. Con base en Muthukrishna et al. (2025) y otros autores, se concluye que el éxito depende de una integración centrada en el ser humano, con infraestructura sólida, formación docente y adaptaciones culturales que garanticen equidad y pertinencia.

Resumo

A inteligência artificial (IA) está remodelando a educação em todo o mundo e a Costa Rica enfrenta o desafio de sua integração. Contudo, o uso inadequado dessas tecnologias pode prejudicar o desenvolvimento educacional. Este ensaio examina os riscos do mau uso da IA em universidades, institutos de línguas e escolas secundárias públicas costarriquenhas. Entre os principais problemas estão a dependência excessiva dos estudantes, o viés algorítmico e a falta de adequação cultural, a desigualdade digital entre regiões urbanas e rurais, a falta de preparação dos professores e as questões éticas de privacidade. Com base em Muthukrishna et al. (2025) e outros pesquisadores, argumenta-se que a IA deve ser integrada dentro de um marco centrado no ser humano, com investimentos em infraestrutura, capacitação docente e adaptações culturais. Só assim poderá contribuir para uma educação mais equitativa e relevante.

The integration of artificial intelligence (AI) into education has generated significant optimism in a country like Costa Rica, a nation that has long emphasized human development as central to its educational policies. From universities in San José to public high schools in rural Guanacaste, AI promises personalized tutoring, automated assessment, and expanded access to knowledge. However, as Muthukrishna, Dai, Panizo Madrid, Sabherwal, Vanoppen, and Yao (2025) caution, “technology alone is not enough” (p. 483). When deployed uncritically or without the necessary systemic support, AI can constrain rather than expand learners’ capabilities, leading to unintended harms that exacerbate existing inequalities and undermine the quality of education students are part of.

One of the most pressing risks is student overdependence on AI tools, which can weaken independent thinking and authentic learning. Muthukrishna, Dai, Panizo Madrid, Sabherwal, Vanoppen, and Yao, H. (2025) note that if AI “spoon-feeds children solutions,” it may erode critical reasoning skills and promote passivity in learners (p. 485). In Costa Rican universities, for instance, students may rely on generative AI to draft essays or complete assignments without engaging deeply with sources and information. Language learners in private institutes often turn to AI translators instead of practicing productive skills, leading to superficial rather than meaningful acquisition and true language practice and production. Darvishi, Khosravi, Sadiq, Gašević, and Siemens (2024) demonstrated that GPT-based tutors can improve short-term performance but simultaneously create dependence, impairing student agency when AI is not available. Long-term mastery is not achieved by overdependent users of AI tutors.

A second major challenge concerns algorithmic bias and cultural mismatch. Because most large language models are trained on Western-centric data, they reproduce values, examples, and idioms that do not align with Costa Rican social and cultural realities. Muthukrishna, Dai, Panizo Madrid, Sabherwal, Vanoppen, and Yao, H. (2025) warn that such bias can “subtly shape values, norms and aspirations” (p. 485) that are alien to the country’s idiosyncrasy. For high school students in Guanacaste or Puntarenas, AI-generated learning materials that reference holidays like Thanksgiving or other contexts unfamiliar to their everyday lives can create disengagement and boredom. Brinkmann, Baumann, Bonnefon, Derex, Müller, Nussberger, and Czaplicka (2023) describe this as the problem of “machine culture,” in which digital systems reinforce dominant cultural narratives at the expense of local knowledge. In Costa Rica, this risk could potentially weaken the role of education in preserving national identity and fostering civic engagement.

The digital divide poses another serious concern in student learning and education. While private universities and international schools in the Central Valley of Costa Rica are well positioned to implement AI, rural schools face persistent deficits in connectivity and access to modern and appropriate digital devices. Muthukrishna, Dai, Panizo Madrid, Sabherwal, Vanoppen, and Yao, H. (2025) emphasize that successful innovations require “reliable electricity, fast internet connectivity, functional and modern computing devices” as preconditions for effective AI integration (p. 487). Without such infrastructure, Costa Rica risks replicating the failures of the One Laptop Per Child initiative, where devices were distributed without systemic support, resulting in little educational gain (UNESCO, 2023). Instead of narrowing learning gaps between private and public education, AI could widen inequalities between urban and rural learners and between students from low-income families and economically advantaged ones.

Equally concerning is the lack of teacher readiness and pedagogical integration. As the article “AI can revolutionise education but technology is not enough: Human development meets cultural evolution” stresses, Estonia and Uruguay succeeded in digital transformation because they invested in teacher training and curricular adaptation, while technology-first approaches failed (Muthukrishna, Dai, Panizo Madrid, Sabherwal, Vanoppen, and Yao, H., 2025). In Costa Rica, if educators are not trained in AI literacy, they may misuse the tools, either delegating too much of their role to technology or rejecting it altogether without giving it a chance. Selwyn (2019) argues that the role of teachers remains irreplaceable: human educators provide the socio-emotional support, cultural interpretation, and moral guidance that AI cannot replicate. Let us always keep in mind that AI has no feelings nor does it care how it is being used for, like cheating for a test. Treating AI as a substitute for teachers could damage the mentoring relationships that are vital in both language learning and adolescent development.

Finally, ethical concerns regarding data privacy and surveillance cannot be ignored when AI is being used. AI systems that track student keystrokes, learning times, or behavior may create a culture of monitoring that “stifles the freedom to fail and learn” (Muthukrishna, Dai, Panizo Madrid, Sabherwal, Vanoppen, and Yao, H., 2025, p. 485). Without clear data governance frameworks, Costa Rican schools in general risk exposing sensitive student information to misuse by private companies. Such practices not only undermine trust but also contradict Costa Rica’s long-standing commitment to human rights and democratic education.

In conclusion, while AI holds transformative potential for Costa Rican education, its wrong use risks undermining student autonomy, deepening inequalities, eroding cultural relevance, weakening teacher roles, and violating privacy. As Muthukrishna, Dai, Panizo Madrid, Sabherwal, Vanoppen, and Yao, H. (2025) argue, successful adoption depends not on the technology itself but on “embedding AI within systems that prioritise infrastructure, teacher training and cultural fit” (p. 490). For Costa Rica, a nation that has historically invested in human-centered development, the lesson must be clear: AI must be guided by ethical safeguards, equitable access policies, and pedagogical strategies that empower rather than displace teachers and students. Otherwise, the promise of AI may become another instance of technological optimism yielding educational disappointment.

📚 References

Brinkmann, L., Baumann, F., Bonnefon, J.-F., Derex, M., Müller, T. F., Nussberger, A.-M., & Czaplicka, A. (2023). Machine culture. Nature Human Behaviour, 7(11), 1855–1868. https://doi.org/10.1038/s41562-023-01742-2

Darvishi, A., Khosravi, H., Sadiq, S., Gašević, D., & Siemens, G. (2024). Impact of AI assistance on student agency. Computers & Education, 210, 104967. https://doi.org/10.1016/j.compedu.2023.104967

Muthukrishna, M., Dai, J., Panizo Madrid, D., Sabherwal, R., Vanoppen, K., & Yao, H. (2025). AI can revolutionise education but technology is not enough: Human development meets cultural evolution. Journal of Human Development and Capabilities, 26(3), 482–492. https://doi.org/10.1080/19452829.2025.2517740 or https://www.tandfonline.com/doi/epdf/10.1080/19452829.2025.2517740?needAccess=true

Selwyn, N. (2019). Should robots replace teachers? AI and the future of education. Polity.

UNESCO. (2023). EdTech tragedy: Lessons from One Laptop per Child. https://www.unesco.org/en/digital-education/ed-tech-tragedy

Potential Policy Brief

Responsible AI Integration in Costa Rican Education: Avoiding Risks, Maximizing Potential

Prepared for: All interested stakeholders
Date: August 2025

Context

Costa Rica is at a crossroads in adopting Artificial Intelligence (AI) in education. Universities, language institutes, and public high schools are experimenting with AI-powered tools for tutoring, grading, and content generation. While these technologies promise personalized learning, increased teacher productivity, and expanded access, misuse or poorly planned implementation risks widening inequalities, eroding student agency, and misaligning with local culture.

Key Risks Identified

1. Overdependence on AI: Students may bypass critical thinking and language production by over-relying on AI-generated answers.

2. Algorithmic Bias & Cultural Mismatch: Foreign-trained AI models may promote content irrelevant to Costa Rican contexts, weakening cultural relevance in learning.

3. Digital Inequality: Unequal infrastructure access between urban private institutions and rural public schools could deepen educational divides.

4. Weak Teacher Training: Without AI literacy and pedagogical integration, teachers may misuse or underuse AI tools.

5. Privacy & Surveillance Concerns: AI platforms collecting student data without transparency risk legal and ethical violations.

6. Technology-First Policies Without Pedagogy: Hardware rollouts without curriculum redesign or teacher support lead to wasted investments.

7. Erosion of Teacher Roles: Cutting human instruction in favor of AI could harm language learning and student motivation.

Recommendations for Costa Rica

1. Adopt a Human-Centred AI Framework:

Define national AI-in-education goals beyond “access to tools” — focus on critical thinking, creativity, and socio-emotional learning.

2. Guarantee Equity of Access

Invest in connectivity, devices, and AI tools for rural and underserved schools.
Support multilingual AI tools and culturally adapted content.

3. Strengthen AI Literacy for Teachers & Students

Integrate AI training into teacher professional development.
Include AI ethics, critical use, and bias awareness in national curricula.

4. Implement Ethical & Privacy Safeguards

Establish clear regulations for data collection, algorithm transparency, and student consent.
Prohibit commercial use of student data collected through educational AI.

5. Co-Design and Pilot Programs

Involve teachers, students, and parents in AI tool selection and adaptation.
Pilot in varied regions before nationwide implementation (“fail locally, learn globally”).

Conclusion
AI can enhance Costa Rica’s educational quality and equity if integrated with infrastructure, teacher empowerment, cultural adaptation, and ethical safeguards. The nation must act deliberately to ensure AI strengthens — not weakens — the capabilities of all learners.

Created by Prof. Jonathan Acuña Solano

August 2025

Based on Muthukrishna, M., Dai, J., Panizo Madrid, D., Sabherwal, R., Vanoppen, K., & Yao, H. (2025). AI can revolutionise education but technology is not enough: Human development meets cultural evolution. Journal of Human Development and Capabilities, 26(3), 482–492. https://doi.org/10.1080/19452829.2025.2517740