About the Guidelines

I am pleased to present these Guidelines for AI in parliaments, which arrive at a crucial moment in our democratic journey. We stand at the threshold of a transformation that is reshaping how parliaments operate and serve their citizens. Artificial Intelligence presents both extraordinary opportunities and significant challenges for our institutions of democracy.

These Guidelines emerge from our recognition that parliaments must take a leading role in governing the use of AI, not only through legislation and oversight but also through their own adoption and implementation of these technologies. The Guidelines represent a collaborative effort, drawing on the expertise and experience of parliamentary staff and technology specialists from across our global community.

The IPU’s commitment to supporting parliaments in their digital transformation journey has never been more vital. We continue to witness first-hand how technological advancement is reshaping parliamentary work. These Guidelines build upon our existing frameworks, including the IPU resolution on “The impact of artificial intelligence on democracy, human rights and the rule of law” and the IPU Charter on Ethics of Science and Technology (October 2024), as well as the Guide to Digital Transformation in Parliaments and the World e-Parliament Reports developed through our Centre for Innovation in Parliament.

The IPU resolution highlights both the tremendous potential of AI to enhance parliamentary functions and the critical importance of ensuring its responsible deployment, particularly in protecting our most vulnerable citizens. These Guidelines therefore emphasise the fundamental principles of transparency, accountability and fairness that must underpin any technological advancement in our institutions of democracy, as well as more technical aspects.

The Guidelines offer a comprehensive framework for parliaments at all stages of their AI journey, whether they are just beginning to explore AI’s potential or are already implementing advanced applications. They provide practical guidance while emphasising the importance of strong governance, ethical considerations and risk management. Most importantly, they stress that AI should augment and enhance human capability rather than replace it, particularly in the context of democratic deliberation and decision-making.

These Guidelines are valuable for different audiences inside and outside of parliament, including for members, especially those serving on modernization committees, technology committees or committees of the future, as well as senior managers and technical experts. They offer useful insights for parliamentarians as they grapple with AI oversight and regulation in society, showing how AI can be deployed responsibly within their own institutions. They provide a framework for making informed decisions about AI adoption while ensuring robust democratic oversight.

Working together, we can ensure that AI serves to strengthen rather than diminish our democratic institutions, upholding the fundamental values that our parliaments represent.

Martin Chungong
Secretary General
Inter-Parliamentary Union

Since the Brazilian Chamber of Deputies’ first experience with artificial intelligence (AI) in 2013, we have been on a journey of continuous learning about this technology and the extraordinary capabilities that it can offer to Parliaments. Despite of that, we are still surprised by the exponential speed of AI advances, and by its ubiquitous nature in the daily lives of public organizations around the world.

Among the lessons learned over the years of using AI in the legislative branch, it is worth to highlight the need for a coordination encompassing many stakeholders, so that its use is well planned and managed. This perception inspired the Parliamentary Data Science Hub meeting in Brasília (in April 2024) to discuss good practices for the use and development of artificial intelligence in parliaments. 

I am proud to note that the excellence of these experts’ work has resulted in a set of guidelines with a friendly approach and flexibility needed to meet the different realities faced by Parliaments. The guidelines combine strategic actions and policies with daily practices, in a way that recognizes the plurality of decisions involved in the use, development and outsourcing of AI systems. 

In addition to being a daily reference document for IT professionals, the “Guidelines for AI in Parliaments” are a reference for directors-general, secretaries-general and managers of Parliaments, in defining strategies and priorities in this topic. I hope that those guidelines will be an instrument for reducing the technological gap between Legislative Houses, so that they can, worldwide, keep up with the advances that society yearns for.

Celso de Barros Correia Neto
Director General
Chamber of Deputies of Brazil

Coordinator of the Parliamentary Data Science hub

Artificial intelligence (AI) presents significant opportunities for parliaments to enhance their operations and to become more efficient and effective, enabling them to better serve citizens. However, adopting AI introduces new challenges and presents risks that must be carefully managed. 

These Guidelines for AI in parliaments (the “Guidelines”) have been developed for parliaments by parliamentary staff and the Inter-Parliamentary Union’s (IPU) Centre for Innovation in Parliament (CIP). They provide comprehensive guidance to support parliaments on their journey towards understanding and implementing AI responsibly and effectively. By adopting a well-thought-through, strategic approach to AI, parliaments can harness the technology’s full potential to drive innovation and efficiency in the legislative process.

The Guidelines cover key areas, including the potential role of AI in parliaments, related risks and challenges, suggested governance structures and AI strategy, ethical principles and risk management, training and capacity-building, and how to manage a portfolio of AI projects across parliament. They are complemented by a set of use cases, shared by parliaments, that describe how AI can support specific parliamentary actions.

The Guidelines stress the importance of a measured, risk-based approach to AI adoption. Key recommendations include the following:

• Start with small pilot projects to build experience.
• Focus on use cases with clear benefits and manageable risks.
• Ensure robust human oversight of AI systems.
• Prioritize transparency and accountability.
• Invest in data and AI literacy across the organization.
• Engage with diverse stakeholders throughout the process.

Audience

The Guidelines have been written to support a range of parliamentary roles:

• For members of parliament (MPs), the Guidelines offer insights into the potential impact of AI on legislative processes, constituent engagement and parliamentary oversight. They provide a clear overview of AI capabilities and limitations, helping MPs to make informed decisions about AI adoption and regulation in parliament.
  
  For members serving on modernisation committees, committees of the future or similar bodies focused on technological advancements, the Guidelines provide strategic insights into AI governance and implementation. This is particularly valuable as parliamentarians increasingly face decisions about AI deployment within their own institutions while simultaneously developing legislation to govern AI use in society more broadly.
   
• For senior parliamentary managers, the Guidelines provide a high-level overview, offering strategic advice on developing AI governance frameworks, policies and oversight mechanisms. This includes establishing clear roles and responsibilities, creating codes of ethics, and aligning AI initiatives with parliament’s organizational goals.
   
• For staff involved in AI implementation, the Guidelines provide detailed guidance on identifying use cases, managing projects, addressing technical challenges, upholding ethical standards and managing risks throughout the AI life cycle.

The Guidelines are designed to support parliaments of all sizes and levels of digital maturity – from large, well-resourced legislatures with advanced digital infrastructures, to smaller parliaments just beginning their digital transformation journey.

The Guidelines can be tailored to allow parliaments to focus on areas most relevant to their current needs and capabilities, and individual parliaments can adapt them to suit their unique circumstances, culture and resources. While digitally mature parliaments may be ready to implement more advanced AI applications, those at earlier stages can use the Guidelines to build foundational governance structures and develop AI literacy. 

Background

In April 2024, the Chamber of Deputies of Brazil – as the host of the CIP’s Parliamentary Data Science Hub – convened, in Brasilia, a panel of experts from parliaments and the IPU to develop an outline of what would become the Guidelines for AI in parliaments. The intention was to develop comprehensive, flexible, scalable guidance to support parliaments as they embrace and responsibly engage with the rapidly emerging technologies associated with AI, whatever their level of digital maturity.

Acknowledgements

The CIP, the Parliamentary Data Science Hub, the IT Governance Hub and the editors are grateful for the contributions of the many parliamentary staff who helped shape, write and review these Guidelines, and for the support provided by the respective parliaments.

Editors

• Patricia Gomes Rêgo de Almeida, Chamber of Deputies of Brazil
• Ludovic Delépine, European Parliament
• Andy Williamson, Centre for Innovation in Parliament, Inter-Parliamentary Union

Contributors

• Patricia Rêgo de Almeida, Chamber of Deputies of Brazil
• Francisco Edmundo Andrade, Chamber of Deputies of Brazil
• Javier de Andrés Blasco, Chamber of Deputies of Spain
• Álvaro Carmo, National Assembly of Angola
• Virginia Carmona, Chamber of Deputies of Chile
• Giovanni Ciccone, Chamber of Deputies of Italy
• Ludovic Delépine, European Parliament
• Claudia di Andrea, Chamber of Deputies of Italy
• Michael Evraire, House of Commons of Canada
• Marcio Fonseca, Chamber of Deputies of Brazil
• José Andrés Jiménez Martín, Chamber of Deputies of Spain
• Vinicius de Morais, Chamber of Deputies of Brazil
• Rune Mortensen, Parliament of Norway
• Neemias Muachendo, National Assembly of Angola
• Jurgens Pieterse, Parliament of South Africa
• Manuel Pereira González, Senate of Spain
• Peter Reichstädter, Parliament of Austria
• Frode Rein, Parliament of Norway
• Esteban Sanchez, Chamber of Deputies of Chile
• Luciana Silo, Chamber of Deputies of Italy
• Paul Vaillancourt, House of Commons of Canada
• Kim van Dooren, Senate of Netherlands
• Marieke van Santen, Senate of Netherlands
• Rodolfo Vaz, Chamber of Deputies of Brazil
• Ricardo Vilarins, Chamber of Deputies of Brazil
• Andy Williamson, Centre for Innovation in Parliament, Inter-Parliamentary Union

Reviewers

• Patricia Rêgo de Almeida, Chamber of Deputies of Brazil
• Ludovic Delépine, European Parliament
• Andy Richardson, Centre for Innovation in Parliament, Inter-Parliamentary Union
• Andy Williamson, Centre for Innovation in Parliament, Inter-Parliamentary Union
• Avinash Bikha, Centre for Innovation in Parliament, Inter-Parliamentary Union 

Graphic design

• Ana Paula A. Diniz, Chamber of Deputies of Brazil
• Kelly Lima Cardoso, Chamber of Deputies of Brazil

Contact

For more information about this work, please contact [email protected]. We are always keen to learn how the Guidelines have been used. We welcome all feedback and suggestions. 

The Guidelines for AI in parliaments are published by the IPU in collaboration with the Parliamentary Data Science Hub in the IPU’s Centre for Innovation in Parliament. This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International licence. It may be freely shared and reused with acknowledgement of the IPU. For more information about the IPU’s work on artificial intelligence, please visit www.ipu.org/AI or contact [email protected].

The role of AI in parliaments | Introducing AI applications | Inter-parliamentary cooperation for AI | Strategic actions towards AI governance | Risks and challenges for parliaments | Generic risks and biases | Ethical principles | Risk management | Alignment with national and international AI frameworks and standards | Project portfolio management | Data governance | Systems development | Security management | Training for Data Literacy and AI Literacy | Glossary of terms