By Elena Pajardi
Introduction
From 2020 onwards, classrooms across Europe transformed almost overnight. The pandemic forced a rapid shift towards digital learning platforms and remote learning environments, accelerating the adoption of educational technologies. AI tools offer real potential: supporting students and teachers through personalised learning, efficient administration and improved online safety. However, these systems also pose serious risks, including reinforcing societal biases through filtering or profiling, compromising privacy and, from the perspective of this article, undermining freedom of expression for both students and teachers.
Although many European countries have developed broad AI strategies, education-specific policies remain uneven. The pandemic placed digital education under a magnifying glass: it showed both how quickly technology could be adopted and how poorly prepared some school systems were to do so responsibly. When monitoring and filtering systems enter classrooms, questions arise: do students still feel free to express themselves? Can teachers use diverse materials without being constrained by algorithms or automated filters?
This article focuses on Italy and Spain to explore the implications of AI-powered monitoring for freedom of expression and consider possible responses. It also draws on international research to illustrate risks that warrant attention in both countries. Not all monitoring and filtering systems use AI, but both conventional and AI-powered systems can affect access to information and opportunities for expression.
Italy: Monitoring, Risks and Educational Rights
In Italy, the rapid adoption of digital tools during the pandemic prompted concerns about student monitoring and privacy. International research provides useful context, although its findings should not be presented as evidence of the prevalence or effects of monitoring in Italian schools.
An Italian article by Carmelina Maurizio, published in Agenda Digitale in 2022, discusses US research by the Center for Democracy & Technology (CDT). The research found that students with learning difficulties or physical disabilities reported greater effects from monitoring. Parents and teachers also expressed concern: approximately 71% of parents and 66% of teachers agreed that students were less likely to share their personal thoughts or ideas openly online if their schools used monitoring systems. Moreover, 66% of teachers feared that students would avoid websites or resources offering support, such as information about coming out or accessing mental health services. These figures relate to the US research discussed in the article, rather than a survey of Italian schools.
The research also highlights a troubling equity dimension. Previous US data cited in the article indicated that approximately six in ten Black students, six in ten Hispanic students, seven in ten rural students and seven in ten students from low-income families relied on school-provided computers or tablets. Students dependent on school-issued devices faced greater exposure to monitoring than peers using personal devices. These findings raise questions relevant to Italy, but further research is needed to establish whether comparable disparities exist in Italian schools.
On the regulatory front, Italy’s 2025 national AI law includes protections for freedom of expression and media pluralism in its provisions on information. Italy is also subject to the EU AI Act, which prohibits the use of AI to infer emotions in educational institutions, except for specified medical or safety purposes. This prohibition has applied since February 2025. Attention monitoring should not automatically be treated as equivalent to prohibited emotion recognition.
However, legal protections must be translated into institutional practice. Educational institutions need to consider how content filtering, device monitoring and remote examination supervision, commonly known as proctoring, can be used without creating an atmosphere of suspicion, self-censorship or restricted speech. The Italian data protection authority’s 2021 action against Bocconi University over its use of proctoring software illustrates the privacy risks associated with educational surveillance.
From the perspective of freedom of expression, the concern is clear: when students and teachers know that their online activity, device use or communications may be watched or analysed by algorithms, they may become less willing to experiment, ask critical questions or explore controversial topics. Teachers may avoid sensitive materials for fear of filters blocking them, while students may hesitate to seek personal support. If monitoring becomes the norm, learning risks turning into compliance.
Spain: Digitisation, Privacy and Expression
Spain faces a similar tension, alongside regional initiatives that offer possible alternatives. The pandemic drove Spanish schools towards online learning, increasing their reliance on digital platforms. Civil society organisations warned that outsourcing educational infrastructure to large technology companies could reduce schools’ control over students’ and teachers’ data and communications.
In 2020, EDRi reported that the Catalan Department of Education had accepted a plan proposed by the digital rights organisation Xnet, titled ‘Privacy and Democratic Digitization of Educational Centers’. The proposal advocated secure servers that respect human rights, auditable tools and training programmes to support a digital culture grounded in rights protection.
The agreement should not be treated as evidence of comprehensive implementation across Catalonia. Nevertheless, subsequent work offers a practical example: Xnet’s DD educational infrastructure brings together free and auditable software tools intended to support schools’ control over data and protect the educational community’s digital rights.
Spain is also subject to the EU AI Act. The Act identifies certain educational applications, including systems used to determine admission or assess learning outcomes, as high-risk uses. These categories should be distinguished from prohibited practices, such as emotion recognition in educational institutions, subject to the limited exceptions described above.
These developments have implications for freedom of expression. Auditable platforms and stronger institutional control over data may help schools explain how information is processed and how restrictions are applied. However, open-source software or public control does not, by itself, guarantee freedom from monitoring or filtering. Transparent policies, appropriate configuration and accessible complaints procedures remain essential.
Where schools use internet filters, classroom management platforms or device monitoring systems, restrictions may prevent access to legitimate resources, particularly when systems are misconfigured or lack transparent review mechanisms. Establishing the extent of these effects in Spanish schools requires evidence about their actual deployment and use.
Effects on Speech, Learning and Rights
The examples from Italy and Spain, considered alongside international research, identify several ways in which monitoring and filtering may affect freedom of expression:
- Self-censorship: Knowing that device use, searches or communications might be monitored can reduce students’ willingness to explore sensitive topics or ask critical questions.
- Restricted access to resources: Overzealous filters may block legitimate academic or personal support content, including mental health information, LGBTQ+ resources and politically or socially sensitive materials.
- Teacher inhibition: Educators may avoid contentious subjects or diverse materials if they believe that algorithms will misinterpret or block them, limiting pedagogical freedom and classroom discussion.
- Unequal impact: Students who depend on school-issued devices may face greater exposure to monitoring. US research documents this concern, but comparable evidence is needed before drawing firm conclusions about its extent in Italy or Spain.
There are also positive developments. Initiatives in Catalonia illustrate how alternative digital infrastructure can support rights protection, while Italy’s AI legislation provides an explicit legal reference point for freedom of expression. However, these examples do not establish that either country has resolved the practical risks associated with educational monitoring.
Policy Recommendations
What can be done to align AI use in schools with freedom of expression? Drawing on the reports and research discussed above, several steps merit consideration:
- Develop education-specific AI policies. General AI strategies should be supported by policies tailored to educational settings, with clear standards for data protection, ethical use and evidence-based decision-making.
- Ensure transparency and participation. Schools should explain which systems they use, what data are collected and how those data are processed. Students and teachers should have meaningful opportunities to shape monitoring policies.
- Adopt minimally intrusive technologies and auditable tools. Prioritise platforms that minimise unnecessary monitoring and allow schools to scrutinise their operation. Open-source tools may support these aims when accompanied by appropriate governance and safeguards.
- Strengthen digital and AI literacy. Educators, students and administrators need to understand how monitoring systems work, what their rights are and what trade-offs are involved. Teacher training should address AI ethics and freedom of expression.
- Implement human oversight, appeals and redress mechanisms. Algorithmic flags should be subject to meaningful human review. Students and teachers should have accessible ways to challenge filtering and monitoring decisions.
- Collect and publish evidence on impacts. European countries should invest in research on how AI affects learning, expression and equality in schools. Country-specific evidence is essential to avoid treating findings from other jurisdictions as proof of local conditions.
- Encourage public–private collaboration with rights safeguards. Where schools work with educational technology providers, agreements should incorporate ethical safeguards, rights-respecting design and public accountability.
By taking these steps, Italy, Spain and other European countries can help ensure that AI’s potential to support learning, improve access and enable innovation does not come at the expense of students’ or teachers’ freedom to express themselves, explore ideas and challenge assumptions. As education becomes increasingly digital, protecting that freedom remains fundamental.
References
Atabey, A. (2025, 20 February). Learning and teaching with AI. Ada Lovelace Institute.
https://www.adalovelaceinstitute.org/blog/learning-and-teaching-with-ai/
Calderini, B. (2020, 24 February). Riconoscimento facciale anche nelle scuole: ecco com’è usato e i problemi di privacy. Agenda Digitale.
Read the article
Center for Democracy & Technology. (2022, 3 August). Hidden harms: The misleading promise of monitoring students online.
Read the report
Council of Europe. (2024). The state of artificial intelligence and education across Europe: Results of a survey of Council of Europe member states.
Publication page
Dimalta, D. (2021, 30 September). Controllo remoto degli studenti, vizio di tanti: il Garante non sanzioni solo Bocconi. Agenda Digitale.
Read the article
European Commission. (2022). Ethical guidelines on the use of artificial intelligence (AI) and data in teaching and learning for educators. Publications Office of the European Union.
https://doi.org/10.2766/153756
European Commission. (n.d.). AI Act.
Policy overview
European Commission. (n.d.). Article 5: Prohibited AI practices. AI Act Service Desk.
Article 5
European Parliament. (2020). The impact of algorithms for online content filtering or moderation: Upload filters (PE 657.101).
Download the study
Italian Republic. (2025, 23 September). Legge 23 settembre 2025, n. 132: Disposizioni e deleghe al Governo in materia di intelligenza artificiale. Gazzetta Ufficiale.
Official legislation
Maurizio, C. (2022, 14 September). Studenti sempre più sorvegliati, non solo a scuola: l’impatto su privacy e salute mentale. Agenda Digitale.
Read the article
UNESCO. (2021). Recommendation on the ethics of artificial intelligence.
UNESCO record
Xnet. (2020, 8 July). Spain: Catalan government agrees to improve privacy in schools. European Digital Rights.
Read the article
Xnet. (n.d.). Introducing DD: A tool for the democratic digitalisation of education.
Project information
