AI in general 1, 2018

Montreal Declaration on Responsible AI



University of Montreal


Well-being, Autonomy, Privacy, Solidarity, Intimacy, Literacy, Bias, Sustainability, Participation, Equity, Diversity, Responsibility,


The Montreal Declaration proposes an ethical framework for the development and deployment of AI. It is meant to guide the digital transition for the benefit of all, and to open a national and international dialogue. Its 10 principles are abstract, non-hierarchical and require appropriate interpretation in particular circumstances. Although they are formulated as ethical principles, they may easily be translated into legal norms. The Declaration is intended for anyone engaging with AI in any capacity, including scientists, technologists and political representatives.
The Declaration came about through a deliberative process that started with a forum in November 2017 followed by a stage of gathering expert and public input through workshops, panels, citizens’ meetings and online commentary. It is to be viewed as an evolving open guidance document and its current state represents the starting point “for an open and inclusive conversation surrounding the future of humanity being served by artificial intelligence technologies.” 
Some of the demands outlined in the Declaration formulate requirements that technological systems or their developers must meet, while others put forward expectations against society, such as digital literacy, which must be supported and enabled by governance mechanisms. 
Members of the public are invited to sign the Declaration online as an expression of their support of its principles. 

  1. Well-being Principle - “the development and use of artificial intelligence systems (AIS) must permit the growth of the well-being of all sentient beings.”

  2. Respect for Autonomy Principle - “AIS must be developed and used while respecting people’s autonomy, and with the goal of increasing people’s control over their lives and their surroundings.” This includes the prohibition of using AIs to suppress certain conceptions of the good life and subjecting people to oppressive regimes of surveillance, classification and incentivization. Furthermore, AIs must not be used to spread misinformation or create confusion between humans and AIs.

  3. Protection of Privacy and Intimacy Principle - introduces the notion of “data acquisition and archiving systems” and states that privacy and intimacy must be protected from intrusion by such systems. Protection extends to personal spaces and to the intimacy of thoughts and emotions. The Declaration states that people’s right to disconnect digitally must be guaranteed. It calls for people’s right to maintain extensive control over information regarding their preferences and to not be subject to influence based on profiling, without consent. In addition, confidentiality and profile anonymity must be guaranteed. Importantly, access to AIS services must not be conditional on giving up control over personal data, and individuals should be free to donate their data for research purposes.

  4. Solidarity Principle - “the development of AIS must be compatible with maintaining the bonds of solidarity among people and generations.” AIS should support human collaboration and foster human relationships. Health care applications must especially consider the patients’ relationships to relatives and the care team. AIS should foster societal conditions with a more equitable distribution of individual and collective risks.

  5. Democratic Participation Principle - “AIS must meet intelligibility, justifiability, and accessibility criteria, and must be subjected to democratic scrutiny, debate, and control.” This Principle includes transparency and understandability requirements for situations when AIS make decisions affecting individuals’ life, quality of life, or reputation. It also calls for the source code of algorithms, public and private, to be available to public authorities and stakeholders for verification. Errors and breaches must be reported to authorities and those affected. The Principle requires that citizens be in a position to deliberate on the objectives and limits of AI systems. Individuals must also always be aware whether they are interacting with an AIS and whether a decision about them has been made by an AIS.

  6. Equity Principle - “The development and use of AIS must contribute to the creation of a just and equitable society.”  AIS must be free of bias, help eliminate relationships of domination between groups of people, and produce social and economic benefits for all. The Principle extends to requiring acceptable working conditions across the entire life cycle of an AIS, from natural resource extraction to data processing. In addition, the activity of AIS users generating data should be recognized as labour. The development and use of ‘commons algorithms’ and open data should be supported.

  7. Diversity Inclusion Principle - “The development and use of AIS must be compatible with maintaining social and cultural diversity and must not restrict the scope of lifestyle choices or personal experiences.” The principle calls for broad inclusion of social diversity in the conceptualisation, development and deployment of algorithms, as well as in their environments of development. AIS must not be used to lock individuals into certain data-derived profiles restricting their personal development or to limit expressions of free expression. AIS offerings must be diversified in each service category to avoid the formation of monopolies.

  8. Prudence Principle - “Every person involved in AI development must exercise caution by anticipating, as far as possible, the adverse consequences of AIS use and by taking the appropriate measures to avoid them.” Especially, mechanisms to monitor dual-use potential must be developed, and open access to applications with a high probability of causing public harms must be restricted. Extensive evaluation measures should be undertaken prior to marketing, and the testing should be open to public authorities and stakeholders.

  9. Responsibility Principle - “The development and use of AIS must not contribute to lessen the responsibility of human beings when decisions must be made.” Humans remain responsible for decisions made by AIS, and in cases where someone’s life, quality of life or reputation are at stake, humans should make informed decisions instead of AIs. Lethal decisions must always be made by humans and can’t be delegated. When harms occur from AIS that are reliable and used as intended, blame should not be placed on its developers. 

  10. Sustainable Development Principle - “The development and use of AIS must be carried out so as to ensure strong environmental sustainability of the planet.” Strong environmental sustainability does not allow the substitution of the loss of natural resources with artificial capital, such as carbon trading. The Principle requires that all digital infrastructure operate at maximum energy efficiency to reduce greenhouse gas emissions, that they generate the least amount of waste, and reduce their impact on the ecosystem and biodiversity from resource extraction to disposal.

AI Governance

This database is an information resource about global governance activities related to artificial intelligence. It is designed as a tool to help researchers, innovators, and policymakers better understand how AI governance is developing around the world.

We're collecting more examples as we find them. If you'd like to add yours to this map, please suggest an entry.

Suggest an entry
  • Filter by:
  • Current filters:
256 entries are listed below
  • Sort by: