Educational Psychology in the age of AI

A Brief Commentary on the APA Report on Artificial Intelligence and Psychology

License Creative Commons

Antoni Hernández-Fernández

Artificial intelligence (AI) is transforming numerous sectors, and psychology is no exception. Indeed, AI has been highlighted as one of the ten key themes in the APA Monitor on Psychology 2025 Trends Report, an annual publication by the American Psychological Association (APA) that warrants careful consideration. The report explores the challenges and opportunities that are likely to shape the future of the discipline. According to the APA, by 2025 AI will not only be reshaping clinical psychology and research but will also be emerging as a key technology within educational psychology. Its anticipated impact includes the personalisation of learning, the analysis of academic behaviour, support for classroom diversity, and the enhancement of teacher training. While it remains uncertain whether such transformations will fully materialise within educational institutions by 2025, one might also hope that AI will ease the bureaucratic burdens that psychologists and educators often face.

The APA asserts that psychologists have a responsibility to actively engage in the ethical design, evaluation, and implementation of these tools. From algorithms that support decision-making to conversational agents that assist learning, the potential is considerable—provided that a critical, ethical, and human-centred approach is upheld, as advocated in the 2017 Barcelona Declaration. What follows is a succinct overview of key elements from the APA report, viewed through the lens of educational application.

Why is AI relevant to educational psychology?

The rise of large language models (such as ChatGPT, DeepSeek, and Gemini) has opened up new possibilities for improving both teaching and learning, as well as addressing mental health challenges among students. Much has already been written about AI’s capacity to personalise learning, support formative assessment, and reduce teacher workload. However, less attention has been paid to its potential in identifying early indicators of academic failure, social exclusion, or emotional distress among pupils.

In this context, the APA implicitly calls upon psychologists to lead the responsible integration of AI into education, guiding decisions based on ethical and scientific standards. This entails ensuring data privacy—particularly critical when working with minors—preserving the centrality of human relationships in educational settings, and promoting equity in access to new technologies, thereby supporting classroom diversity and inclusion.

Potential applications 

The APA has outlined a range of key resources, summarised in Table 1, designed to assist in the training of professionals in both psychology and education. One of the most promising areas is the use of AI for formative feedback and learning supervision. For instance, intelligent tutoring systems can analyse student responses and provide instant feedback, adjusting the difficulty level or suggesting reinforcement activities.

When applied judiciously, AI can be an ally in reducing educational inequalities. Generative models make it possible to adapt educational materials to diverse cultural contexts more easily than before—not merely by translating content, but by explaining social norms and relevant references tailored to different groups. This can significantly enhance inclusion and wellbeing for pupils with migrant backgrounds, linguistic minorities, or special educational needs. Another promising application—which I often highlight in training sessions—is the use of AI to draft materials and activities aligned with principles of clear communication, which can be particularly beneficial for students with reading comprehension difficulties.

In line with this communicative improvement, AI tools are being developed to assist teachers in transcribing, analysing, and refining their practice using data collected during lessons or meetings with the form tutor. These tools might identify patterns in student participation or group emotional dynamics, generating alerts for early intervention. Furthermore, the APA notes that AI may help expand access to mental health resources for students unable to attend in-person therapy, through chatbots trained to detect emotional states or deliver psychological first aid. Such tools, however, must always operate in close collaboration with human professionals who are present and accessible in the learner’s environment, and under appropriate ethical supervision.

In the domain of professional development, AI can also enrich the initial and ongoing training of educators and psychologists via simulations and virtual environments in which to practise guidance, mediation, or teaching skills. Just as in clinical contexts, models trained to detect empathy or active listening may be used to support the development of interpersonal communication skills, allowing trainees to rehearse with machines before entering real-world scenarios.

Table 1. Key Resources on AI and Psychology (APA)

Source: Adapted and expanded from the American Psychological Association (APA, 2025), APA Monitor on Psychology 2025 Trends Report.

A critical evaluation—and a call to action 

Nevertheless, the APA’s report invokes the precautionary principle. Beyond the well-documented environmental and social costs of generative AI (as illustrated in the AI Cartography by the Estampa collective), many AI applications in psychology remain underdeveloped and pose real risks: threats to privacy, data bias, technological dependency, and the dehumanisation of educational relationships, among the most pressing concerns. Educational psychology thus bears a responsibility to critically examine these tools and define reasonable boundaries for their use.

This includes, for instance, questioning who has access to student data, whether such data are being used to train proprietary models by corporations (e.g. OpenAI, Google, Microsoft), whether consent can be revoked, and how automated decisions are explained. It is equally vital to reflect on how teaching is affected when certain tasks are delegated to machines—what impact this has on motivation, pedagogical reflection, and relationships with students. A promising approach involves the use of open-source AI models that can be run locally, providing full control over data and goals, rather than relying on large commercial platforms that often operate beyond EU regulations and offer little transparency regarding training data and data destinations.

AI is not a neutral tool. Its impact on education will largely depend on how it is implemented. For this reason, educational psychology must not merely adopt AI tools, but actively participate in their design, regulation, and evaluation from a techno-ethical, critical, and humanistic perspective. As the APA report affirms: the revolution is already upon us. Rather than fearing it, now is the time to engage with it critically, to examine how it may support educators and students—yet not at any cost. One of the foremost challenges for current research lies in examining the empirical evidence on whether AI genuinely enhances learning, equity, and the holistic development of individuals.

___

References:

APA (2025). APA Monitor on Psychology 2025. APA 2025 Ten Emerging Trends in Psychology | Free Report

Source: educational EVIDENCE

Rights: Creative Commons