The Guidelines for the Responsible Use of Artificial Intelligence Tools at the University were approved by Resolution No. 153/2026 of the Board of Directors on 22 May 2026. The official document is available on the University portal, in the “Statute and Regulations” menù, “Regolamenti Sistemi Informativi” section.
The text of the Guidelines is provided below in a navigable format, complemented by examples and use cases.
Introduction and Objectives
Artificial Intelligence (AI), and generative AI in particular, represents a resource with the potential to transform the way we study, teach, conduct research, and work. The University of Pavia recognizes this technology as an opportunity to strengthen the competitiveness of its education and research system while preserving quality and promoting fairness.
The purpose of this document is to provide a framework for the responsible, ethical, and legally compliant use of Artificial Intelligence tools in support of all University activities, while safeguarding the fundamental values of quality, academic integrity, and the protection of individuals. This document is intended to be dynamic and may be updated and supplemented over time in response to technological and regulatory developments. All members of the University community are expected to comply with the principles set out herein.
The University promotes the use of AI tools, giving preference, where available, to solutions integrated within institutional environments (such as collaborative platforms officially adopted by the University), in order to ensure higher levels of security, control, and regulatory compliance.
General Principles
Human-Centred Approach: AI is a tool intended to support activities and must never replace the central role of human beings in carrying them out. Continuous human oversight must be ensured for all systems used within the University, including the evaluation, validation, and, where necessary, revision and correction of outputs before their use.
Responsibility: Anyone choosing to use Artificial Intelligence tools is personally responsible for their use and for critically assessing the accuracy of generated outputs. Particular attention should be paid to the risk of inaccurate, unverifiable, or misleading information, as well as to the variability of outputs produced by AI systems.
The use of Artificial Intelligence must comply with the principle of “do no significant harm”, avoiding applications that may negatively affect individuals, communities, fundamental rights, safety, fairness, or the environment.
The use of tools other than those directly provided by the University, including tools acquired through personal or project funds, entails an additional responsibility on the part of the user to assess their characteristics, reliability, legal compliance (particularly with regard to data protection, security, and intellectual property), and suitability for the intended context of use.
Transparency: Any significant use of Artificial Intelligence (i.e., beyond the mere collection and organization of information), whether in teaching, research, or administrative activities, must be explicitly disclosed, specifying the tool used, the scope of application, and the manner in which it was employed.
Academic Integrity: The use of Artificial Intelligence must comply with the principles of academic integrity, ensuring transparency, accuracy, and respect for intellectual property in learning, research, and assessment processes. AI tools may support individual activities but must not replace personal contribution, critical thinking, or intellectual responsibility.
Security, Data Protection, and Confidentiality: The use of Artificial Intelligence tools must comply fully with applicable regulations concerning personal data protection and confidentiality. Any use of AI systems involving the processing of personal data must be assessed in advance in accordance with Regulation (EU) 2016/679 (GDPR) and in line with the regulations, guidelines, and organizational frameworks adopted by the University.
It is prohibited to enter special categories of personal data, confidential information, or strategic information (such as internal documents, information subject to confidentiality obligations, or content protected by intellectual property rights) into AI systems that have not undergone prior assessment and approval by the University.
Users must also ensure appropriate levels of security when using AI tools, adopting measures to prevent unauthorized access and protecting their accounts, for example through multi-factor authentication.
Training and AI Literacy: In order to promote the responsible and informed use of Artificial Intelligence tools, the University is committed to providing training opportunities for faculty, students, and professional services staff. Such training aims to develop the skills needed to understand how AI tools function, critically evaluate their outputs, adopt good practices, and gain awareness of the limitations and risks associated with Artificial Intelligence.
Inclusiveness and Environmental Sustainability: Artificial Intelligence tools must be used in accordance with the principles of inclusiveness, fairness, and sustainability. Members of the University community who use AI tools are responsible for critically evaluating the outputs generated in order to prevent or mitigate discriminatory effects on individuals or groups, particularly those in vulnerable situations. Ensuring the quality and appropriateness of data used throughout the development and deployment of AI systems is essential.
Given the significant energy consumption associated with many Artificial Intelligence solutions, their use should also be assessed from an environmental perspective, favouring models and infrastructures that are proportionate to the objectives pursued.
Guidelines for Teaching
The University encourages the responsible use of Artificial Intelligence tools in teaching activities, provided that such tools support rather than replace teaching and learning processes. Their use must respect academic freedom and promote fair, transparent, and responsible approaches to learning. AI tools must always be used under human supervision and with appropriate critical evaluation of the outputs generated.
In order to safeguard the integrity of teaching and learning processes, uses of Artificial Intelligence that improperly replace students’ personal work, compromise the fairness, transparency, or integrity of assessment processes, violate data protection regulations, or negatively affect the quality of teaching are prohibited.
For Faculty Members and Language Instructors
Autonomy: Faculty members retain autonomy in determining AI usage policies within their courses, including the authorization or prohibition of specific tools.
Communication: Rules governing the use of AI in both course activities and final assessments must be clearly stated in the course syllabus at the beginning of the academic year and adequately communicated to students during teaching activities.
Promoting Awareness: Faculty members are encouraged to foster students’ understanding of the importance of responsibility, transparency, integrity, and fairness in the use of AI, thereby supporting ethical academic conduct and the development of individual competencies.
Assessment: The assessment of student learning outcomes, particularly in activities contributing to final grades, must never be entirely delegated to AI tools.
Course Design: Faculty members are encouraged to design examinations and assessment activities while taking into account the possible use of AI tools and adapting evaluation methods accordingly.
Content Generation: Faculty members may independently determine the contexts in which AI may be useful for generating teaching materials, educational content, or examination questions, while remaining fully compliant with the General Principles outlined in these Guidelines.
Learning Support: AI tools may be developed and made available to students as learning support resources, study aids, and instruments for enhancing student assistance, without replacing the tutoring and support services provided by the University.
Theses and Dissertations: Faculty members and/or academic boards may introduce additional requirements or restrictions concerning the use of AI in the preparation of theses and dissertations. Such requirements must be communicated appropriately to students.
For Students
Learning Support: AI may be used as a tool to enhance learning, for example through brainstorming, idea organization, language revision, and grammar correction of texts independently produced by students. Artificial Intelligence must never replace students’ own critical thinking and personal intellectual contribution.
Critical Use: Students are expected to use AI tools critically, verifying the accuracy and reliability of generated information and assuming responsibility for the content they produce.
Responsibility: Students remain fully responsible for all work submitted as their own, including work produced with the support of AI tools.
Compliance with Teaching Requirements: The use of AI must comply with the instructions provided by faculty members for each course, assessment activity, or programme of study.
Limits on Use: The use of AI to generate answers during examinations or assessments is prohibited unless explicitly authorized by the instructor. Likewise, AI may not be used for the complete drafting of dissertations, theses, or academic assignments without substantial personal critical revision and contribution. Any violations may be assessed by the competent disciplinary bodies.
Mandatory Disclosure: Any assignment, dissertation, or thesis prepared with the assistance of AI tools must include a statement specifying the tools used, the scope of their use, and the manner in which they were employed.
Guidelines for Research
Research Integrity: Researchers remain solely responsible for methodological quality, data interpretation, the accuracy of results, and compliance with ethical standards and good research practices. The use of Artificial Intelligence tools must always take place under human supervision. Researchers are required to verify the correctness of generated outputs, including data, analyses, code, and bibliographic references, and to adequately document the procedures adopted.
Authorship: An Artificial Intelligence system cannot be considered an author or co-author of a scientific publication. Authors remain fully responsible for the content produced, including content developed with the support of AI tools.
Transparency: The use of AI tools in scientific production must be disclosed transparently, specifying the methods, tools, and scope of use, in accordance with the editorial policies of the relevant publication venues.
Data Use and Confidentiality: The University promotes the use of Artificial Intelligence to support scientific research and technological innovation. Researchers remain responsible for ensuring that such use complies with applicable legislation and regulations.
Guidelines for Technical, Administrative and Managerial Activities
This section complements the General Principles of the University Guidelines concerning the use of generative Artificial Intelligence tools in technical, administrative, and managerial activities, with reference to the University’s role as a deployer of AI systems and to all individuals involved, in various capacities, in the related processes, including institutional communication activities.
Service Improvement: Artificial Intelligence is encouraged as a means of supporting staff in repetitive or time-consuming tasks.
Human Oversight: Any act, document, or output produced with the support of Artificial Intelligence must be reviewed and validated by the responsible staff member. The use of AI systems for fully automated decision-making without appropriate human intervention is prohibited.
Data Protection: Sensitive, confidential, or restricted University information must not be entered into Artificial Intelligence tools that have not been previously assessed, certified, or approved by the University. The use of personal data is permitted only through authorized tools and in compliance with the principles of data minimization, security, and processing control.
Responsibility in the Use of Tools: Staff members are responsible for the use of Artificial Intelligence tools and for verifying generated outputs before their use or dissemination. The use of unauthorized tools is undertaken entirely at the user’s own responsibility.
Transparency: The use of Artificial Intelligence in administrative processes must be communicated transparently. Depending on the significance of the process in which AI tools are employed, the University undertakes to provide a clear and meaningful explanation of the role played by the system in the decision-making process.
Automated Systems and Agents: Non-conversational and semi-autonomous AI tools (also referred to as “agents”), which may be used to automate certain administrative processes, must include clear mechanisms allowing a responsible operator to supervise, interrupt, and control ongoing activities. Prior to deployment, the behaviour of such agents must be appropriately tested in a controlled environment.
Monitoring: The use of Artificial Intelligence must be subject to continuous monitoring in order to assess effectiveness, identify potential issues, and update usage practices where necessary.
The University promotes the use of institutional or authorized AI tools that have undergone prior technical, legal, and information security assessment, including evaluation of contractual conditions and data processing arrangements. The use of AI tools that are not included within the University’s approved portfolio is strongly discouraged unless specifically assessed and approved in advance.
Staff members remain responsible for their use of AI tools and for verifying generated outputs. Any use of unauthorized tools is undertaken entirely at the responsibility of the individual user.
The use of Artificial Intelligence tools within the University’s administrative processes must be agreed upon within the activities and governance structures of the relevant administrative units.
Use Cases and Practical Examples
The Guidelines include practical examples and use cases that form an integral part of the document. To facilitate consultation and navigation, a dedicated page containing detailed use cases and practical examples has been made available here.
Legal References and University Regulations
This section brings together the main national and European legal references, as well as the University’s internal regulations, that are relevant to the use of Artificial Intelligence tools within institutional activities. These Guidelines are intended to complement and build upon these sources by providing operational guidance and practical recommendations, including for specific use cases, while fully respecting the validity and precedence of applicable laws and regulations. University regulations, including the organizational and security frameworks adopted by the University in implementation of the NIS Directive (EU 2022/2555), may further specify and elaborate on the provisions contained in this document.
University Regulations and Internal Sources:
University General Regulations
Code of Ethics
Academic Regulations
Student Regulations
Regulations on Personal Data Processing and the Use of IT Resources and Communication Services
Regulations of individual Departments and Degree Programmes.
National and European Legal References:
European Artificial Intelligence Regulation (AI Act)
Regulation (EU) 2016/679 (GDPR)
Directive (EU) 2022/2555 (NIS2)
Italian Privacy Code and Related Legislation
Legislative Decree No. 82 of 7 March 2005 (Digital Administration Code – CAD)
Italian Copyright Law (Law No. 633 of 22 April 1941)
Law No. 132/2025 – Provisions and Delegations to the Government on Artificial Intelligence
EU ERA Living Guidelines for the Responsible Use of Generative AI in Research