The guidelines on AI 2025 | Vatican

January 2025

View the original document

Strategies and Action Plans

The guidelines established by the Pontifical Commission for the State of Vatican City represent a forward-thinking approach to integrating artificial intelligence (AI) into modern society while adhering to ethical and moral principles deeply rooted in Catholic teachings. Recognising the dual potential of AI—as both a powerful tool for human progress and a possible source of harm to individuals and communities—the Vatican seeks to ensure that these technologies are developed and used responsibly, with a firm focus on the dignity of every human being and the common good.

This framework emerges in response to the rapid evolution of AI systems and their increasingly pervasive influence on all aspects of life, from healthcare and education to governance and cultural heritage. The guidelines aim to address the ethical dilemmas posed by these advancements, emphasising the need for transparency, accountability, and a human-centred approach in AI applications. By doing so, the Vatican underscores the belief that technological progress should not come at the expense of moral values or the well-being of individuals, especially the most vulnerable members of society.

Through these principles, the Vatican aims to position itself as a moral and ethical steward in the global discourse on AI, calling for a balance between innovation and responsibility. The document highlights that AI is a gift of human creativity, which, in turn, is seen as a reflection of divine grace. However, this gift must be safeguarded against misuse, particularly practices that undermine human autonomy, create social inequalities, or conflict with the Church’s mission.

The guidelines are designed to apply within the Vatican’s governance structure and its judicial and administrative bodies and extend to economic operators, contractors, and other entities engaging with AI in the Vatican’s jurisdiction. By articulating these principles, the Vatican aims to lead by example, setting a standard for ethical AI governance that aligns with its spiritual and humanitarian values.

General provisions

The guidelines aim to ensure the ethical and transparent use of AI in a way that respects human dignity and serves the common good. These principles apply within Vatican City’s territory and its associated operations.

Key definitions

AI is defined as systems capable of data analysis, learning, decision-making, and performing tasks typically requiring human intelligence. The document also defines related terms such as data, biometric data, and AI systems.

Fundamental principles

Ethical responsibility: Activities involving AI must prioritise human dignity, common good, ethical accountability, and transparency.
Data security: Systems must ensure data privacy, prevent discrimination, and support sustainable practices.
Anthropocentric design: AI should serve humanity, respecting human autonomy and decision-making.

Prohibited uses

The guidelines ban:

Discriminatory practices based on anthropological deductions.
Manipulative techniques causing harm.
Exclusion of individuals with disabilities.
Social inequalities or violations of human dignity.
Threats to Vatican security or public order.
Uses conflict with the mission of the Church or the Pope.

Specific application areas

Data and information: AI must ensure accurate, unbiased data use, respecting personal freedom and privacy.
Scientific research and healthcare: AI can aid public health and medical care, but human rights and privacy must be preserved.
Intellectual property: AI-generated content must respect copyright laws, and its origin should be marked.
Cultural heritage: AI should enhance preservation without compromising the integrity of cultural artifacts.
Infrastructure and services: AI must promote sustainability and safety without hindering decision-making processes.
Administrative procedures: AI should streamline administration while maintaining transparency and human oversight.
Workplace practices: AI may enhance training and safety but must avoid discrimination.
Judiciary: AI can support legal research and organisation but cannot influence judicial decisions.
Security: AI applications in cybersecurity and safety will follow future regulations.

Implementation and oversight

AI commission: A five-member commission, chaired by the Secretary General, will oversee AI-related developments, monitor risks, and evaluate impacts through biannual reports.
Future legislation: Additional laws and regulations will be developed within a year of the enforcement of these guidelines.