OECD AI Principles

The OECD AI Principles promote use of AI that is innovative and trustworthy and that respects human
rights and democratic values. Adopted in May 2019, they set standards for AI that are practical and
flexible enough to stand the test of time.

Contents
Contents .............................................................................................................. 1
Principles for trustworthy AI................................................................................. 2
Values-based principles ......................................................................................... 3
Inclusive growth, sustainable development and well-being (Principle 1.1) .............. 3
Rationale for this principle .............................................................................. 3
Respect for the rule of law, human rights and democratic values, including fairness
and privacy (Principle 1.2) ................................................................................. 4
Rationale for this principle .............................................................................. 4
Transparency and explainability (Principle 1.3) ................................................... 4
Rationale for this principle .............................................................................. 5
Robustness, security and safety (Principle 1.4) .................................................... 6
Rationale for this principle .............................................................................. 6
Accountability (Principle 1.5) ............................................................................. 7
Rationale for this principle .............................................................................. 7
Recommendations for policymakers....................................................................... 8
Investing in AI research and development (Principle 2.1) ..................................... 8
Rationale for this principle .............................................................................. 8
Fostering an including AI-enabling ecosystem (Principle 2.2)................................ 9
Rationale for this principle .............................................................................. 9
Shaping and enabling interoperable governance and policy environment for AI
(Principle 2.3) ................................................................................................. 10
Rationale for this principle ............................................................................ 10
Building human capacity and preparing for labour market transformation
(Principle 2.4) ................................................................................................. 11

OECD AI Principles | 1
 Rationale for this principle ............................................................................ 11
International co-operation for trustworthy AI (Principle 2.5) ............................. 12
Rationale for this principle ............................................................................ 12

Principles for trustworthy AI

The OECD AI Principles were initially adopted in 2019 and updated in May 2024. Adherents updated
them to consider new technological and policy developments, ensuring they remain robust and fit for
purpose.

The Principles guide AI actors in their efforts to develop trustworthy AI and provide policymakers
with recommendations for effective AI policies.

Countries use the OECD AI Principles and related tools to shape policies and create AI risk
frameworks, building a foundation for global interoperability between jurisdictions. Today,
the European Union, the Council of Europe, the United States, and the United Nations and other
jurisdictions use the OECD’s definition of an AI system and lifecycle below in their legislative and
regulatory frameworks and guidance. The principles, definition and lifecycle are all part of the OECD
Recommendation on Artificial Intelligence.

OECD AI Principles | 2
Values-based principles

Inclusive growth, sustainable development and well-being (Principle 1.1)

Stakeholders should proactively engage in responsible stewardship of trustworthy AI in
pursuit of beneficial outcomes for people and the planet, such as augmenting human
capabilities and enhancing creativity, advancing inclusion of underrepresented
populations, reducing economic, social, gender and other inequalities, and protecting
natural environments, thus invigorating inclusive growth, well-being, sustainable
development and environmental sustainability.

Rationale for this principle

This principle recognises that guiding the development and use of AI toward prosperity and beneficial
outcomes for people and planet is a priority. Trustworthy AI can play an important role in advancing
inclusive growth, sustainable development and well-being and global development objectives.
Indeed, AI can be leveraged for social good and can substantially contribute to achieving the
Sustainable Development Goals (SDGs) in areas such as education, health, transport, agriculture,
environment, and sustainable cities, among others.

This stewardship role should strive to address concerns about inequality and the risk that disparities
in technology access increase existing divides within and between developed and developing
countries. The OECD Framework for Policy Action on Inclusive Growth provides a useful point of
reference. This framework is intended to guide policy action that “brings everyone along” towards a
more robust, confident future.

This principle also recognises that AI systems could perpetuate existing biases and have a disparate
impact on vulnerable and underrepresented populations, such as ethnic minorities, women, children,
the elderly and the less educated or low skilled. Disparate impact is a particular risk in low- and
middle-income countries. This principle emphasises that AI can also, and should, be used to empower
all members of society and to help reduce biases.

Responsible stewardship is furthermore a recognition that throughout the AI system lifecycle, AI

actors and stakeholders can, and should, encourage the development and deployment of AI for
beneficial outcomes with appropriate safeguards. Defining these beneficial outcomes, and how best
to achieve them will benefit from multidisciplinary and multi-stakeholder collaboration and social
dialogue. Furthermore, a meaningful, well-informed, and iterative public dialogue that is inclusive of
all stakeholders can enhance public trust in and understanding of AI.

1 Referred to in the preamble of the Recommendation.

2OECD (2018), Opportunities for All: A Framework for Policy Action on Inclusive Growth, OECD Publishing, Paris,
[Link]

OECD AI Principles | 3
Respect for the rule of law, human rights and democratic values, including
fairness and privacy (Principle 1.2)
AI actors should respect the rule of law, human rights, democratic and human-centred
values throughout the AI system lifecycle. These include non-discrimination and equality,
freedom, dignity, autonomy of individuals, privacy and data protection, diversity,
fairness, social justice, and internationally recognised labour rights. This also includes
addressing misinformation and disinformation amplified by AI, while respecting freedom
of expression and other rights and freedoms protected by applicable international law.
To this end, AI actors should implement mechanisms and safeguards, such as capacity
for human agency and oversight, including to address risks arising from uses outside of
intended purpose, intentional misuse, or unintentional misuse in a manner appropriate
to the context and consistent with the state of the art.

Rationale for this principle

AI should be developed consistent with human-centred values, such as fundamental freedoms,
equality, fairness, rule of law, social justice, data protection and privacy, as well as consumer rights
and commercial fairness.

Some applications or uses of AI systems have implications for human rights, including risks that
human rights (as defined in the Universal Declaration of Human Rights) 1 and human-centred values
might be deliberately or accidently infringed. It is therefore important to promote “values-alignment”
in AI systems (i.e., their design with appropriate safeguards) including capacity for human
intervention and oversight, as appropriate to the context. This alignment can help ensure that AI
systems’ behaviours protect and promote human rights and align with human-centred values
throughout their operation. Remaining true to shared democratic values will help strengthen public
trust in AI and support the use of AI to protect human rights and reduce discrimination or other unfair
and/or unequal outcomes.

This principles also acknowledges the role of measures such as human rights impact assessments
(HRIAs) and human rights due diligence, human determination (i.e., a “human in the loop”), codes
of ethical conduct, or quality labels and certifications intended to promote human-centred values and
fairness.

1 Available at: [Link]

Transparency and explainability (Principle 1.3)

AI Actors should commit to transparency and responsible disclosure regarding AI
systems. To this end, they should provide meaningful information, appropriate to the
context, and consistent with the state of art:
• to foster a general understanding of AI systems, including their capabilities and
limitations,
• to make stakeholders aware of their interactions with AI systems, including in the
workplace,

OECD AI Principles | 4
 • where feasible and useful, to provide plain and easy-to-understand information on
the sources of data/input, factors, processes and/or logic that led to the prediction,
content, recommendation or decision, to enable those affected by an AI system to
understand the output, and,
• to provide information that enable those adversely affected by an AI system to
challenge its output.

Rationale for this principle

The term transparency carries multiple meanings. In the context of this Principle, the focus is first on
disclosing when AI is being used (in a prediction, recommendation or decision, or that the user is
interacting directly with an AI-powered agent, such as a chatbot). Disclosure should be made with
proportion to the importance of the interaction. The growing ubiquity of AI applications may
influence the desirability, effectiveness or feasibility of disclosure in some cases.

Transparency further means enabling people to understand how an AI system is developed, trained,
operates, and deployed in the relevant application domain, so that consumers, for example, can make
more informed choices. Transparency also refers to the ability to provide meaningful information and
clarity about what information is provided and why. Thus transparency does not in general extend to
the disclosure of the source or other proprietary code or sharing of proprietary datasets, all of which
may be too technically complex to be feasible or useful to understanding an outcome. Source code
and datasets may also be subject to intellectual property, including trade secrets.

An additional aspect of transparency concerns facilitating public, multi-stakeholder discourse and the
establishment of dedicated entities, as necessary, to foster general awareness and understanding of AI
systems and increase acceptance and trust.

Explainability means enabling people affected by the outcome of an AI system to understand how it
was arrived at. This entails providing easy-to-understand information to people affected by an AI
system’s outcome that can enable those adversely affected to challenge the outcome, notably – to the
extent practicable – the factors and logic that led to an outcome. Notwithstanding, explainability can
be achieved in different ways depending on the context (such as, the significance of the outcomes).
For example, for some types of AI systems, requiring explainability may negatively affect the
accuracy and performance of the system (as it may require reducing the solution variables to a set
small enough that humans can understand, which could be suboptimal in complex, high-dimensional
problems), or privacy and security. It may also increase complexity and costs, potentially putting AI
actors that are SMEs at a disproportionate disadvantage.

Therefore, when AI actors provide an explanation of an outcome, they may consider providing – in
clear and simple terms, and as appropriate to the context – the main factors in a decision, the
determinant factors, the data, logic or algorithm behind the specific outcome, or explaining why
similar-looking circumstances generated a different outcome. This should be done in a way that
allows individuals to understand and challenge the outcome while respecting personal data protection
obligations, if relevant.

OECD AI Principles | 5
Robustness, security and safety (Principle 1.4)
AI systems should be robust, secure and safe throughout their entire lifecycle so that, in
conditions of normal use, foreseeable use or misuse, or other adverse conditions, they
function appropriately and do not pose unreasonable safety and/or security risks.
Mechanisms should be in place, as appropriate, to ensure that if AI systems risk causing
undue harm or exhibit undesired behaviour, they can be overridden, repaired, and/or
decommissioned safely as needed.
Mechanisms should also, where technically feasible, be in place to bolster information
integrity while ensuring respect for freedom of expression.

Rationale for this principle

Addressing the safety and security challenges of complex AI systems is critical to fostering trust in
AI. In this context, robustness signifies the ability to withstand or overcome adverse conditions,
including digital security risks. This principle further states that AI systems should not pose
unreasonable safety risks including to physical security, in conditions of normal or foreseeable use or
misuse throughout their lifecycle. Existing laws and regulations in areas such as consumer protection
already identify what constitutes unreasonable safety risks. Governments, in consultation with
stakeholders, must determine to what extent they apply to AI systems.

AI actors can employ a risk management approach (see below) to identify and protect against
foreseeable misuse, as well as against risks associated with use of AI systems for purposes other than
those for which they were originally designed. Issues of robustness, security and safety of AI are
interlinked. For example, digital security can affect the safety of connected products such as
automobiles and home appliances if risks are not appropriately managed.

The Recommendation highlights two ways to maintain robust, safe and secure AI systems:

1. traceability and subsequent analysis and inquiry, and

2. applying a risk management approach.
like explainability (see 1.3), traceability can help analysis and inquiry into the outcomes of an AI
system and is a way to promote accountability. Traceability differs from explainability in that the
focus is on maintaining records of data characteristics, such as metadata, data sources and data
cleaning, but not necessarily the data themselves. In this, traceability can help to understand
outcomes, to prevent future mistakes, and to improve the trustworthiness of the AI system.

Risk management approach: The Recommendation recognises the potential risks that AI systems pose
to human rights, bodily integrity, privacy, fairness, equality and robustness. It further recognises the
costs of protecting from these risks, including by building transparency, accountability, safety and
security into AI systems. It also recognises that different uses of AI present different risks, and some
risks require a higher standard of prevention or mitigation than others.

A risk management approach, applied throughout the AI system lifecycle, can help to identify, assess,
prioritise and mitigate potential risks that can adversely affect a system’s behaviour and outcomes.
Other OECD standards on risk management, for example in the context of digital security risk

OECD AI Principles | 6
management and risk-based due diligence under the MNE Guidelines and OECD Due Diligence
Guidance for Responsible Business Conduct, may offer useful guidance 1. Documenting risk
management decisions made at each lifecycle phase can contribute to the implementation of the other
principles of transparency (1.3) and accountability (1.5).

1[Link]

Accountability (Principle 1.5)

AI actors should be accountable for the proper functioning of AI systems and for the
respect of the above principles, based on their roles, the context, and consistent with the
state of the art.
To this end, AI actors should ensure traceability, including in relation to datasets,
processes and decisions made during the AI system lifecycle, to enable analysis of the AI
system’s outputs and responses to inquiry, appropriate to the context and consistent with
the state of the art.
AI actors, should, based on their roles, the context, and their ability to act, apply a
systematic risk management approach to each phase of the AI system lifecycle on an
ongoing basis and adopt responsible business conduct to address risks related to AI
systems, including, as appropriate, via co-operation between different AI actors,
suppliers of AI knowledge and AI resources, AI system users, and other stakeholders.
Risks include those related to harmful bias, human rights including safety, security, and
privacy, as well as labour and intellectual property rights.

Rationale for this principle

The terms accountability, responsibility and liability are closely related yet different, and also carry
different meanings across cultures and languages. Generally speaking, “accountability” implies an
ethical, moral, or other expectation (e.g., as set out in management practices or codes of conduct) that
guides individuals’ or organisations’ actions or conduct and allows them to explain reasons for which
decisions and actions were taken. In the case of a negative outcome, it also implies taking action to
ensure a better outcome in the future. “Liability” generally refers to adverse legal implications arising
from a person’s (or an organisation’s) actions or inaction. “Responsibility” can also have ethical or
moral expectations and can be used in both legal and non-legal contexts to refer to a causal link
between an actor and an outcome.

Given these meanings, the term “accountability” best captures the essence of this principle. In this
context, “accountability” refers to the expectation that organisations or individuals will ensure the
proper functioning, throughout their lifecycle, of the AI systems that they design, develop, operate or
deploy, in accordance with their roles and applicable regulatory frameworks, and for demonstrating
this through their actions and decision-making process (for example, by providing documentation on
key decisions throughout the AI system lifecycle or conducting or allowing auditing where justified).

OECD AI Principles | 7
Recommendations for policymakers

Investing in AI research and development (Principle 2.1)

Governments should consider long-term public investment, and encourage private
investment, in research and development and open science, including interdisciplinary
efforts, to spur innovation in trustworthy AI that focus on challenging technical issues
and on AI-related social, legal and ethical implications and policy issues.
Governments should also consider public investment and encourage private investment
in open-source tools and open datasets that are representative and respect privacy and
data protection to support an environment for AI research and development that is free
of harmful bias and to improve interoperability and use of standards.

Rationale for this principle

Scientific breakthroughs enabled by AI could help solve societal challenges and create entirely new
industries. These possibilities underscore the importance of basic research and of considering long
time horizons in research policy. While the private sector has taken the lead in applied AI R&D
investments in recent years, governments, at times complemented by foundations focused on public
good, have an important role to play in providing sustained investment in public research with long-
term horizons. This type of investment is essential to driving and shaping trustworthy AI innovation
and ensuring beneficial outcomes for all, particularly areas under-served by market-driven
investments. Publicly funded research can help address challenging technological issues that affect a
broad range of AI actors and stakeholders. AI research includes research in: AI applications, such as
natural language processing; techniques to teach AI systems, such as neural networks; optimisation
notably to reduce the amount of data required for AI development, such as one-shot-learning; and
research addressing societal considerations, such as transparency and explainability, as well as
technologies to protect data integrity.

In addition, because AI has broad reach and pervasive implications on multiple facets of life, this
recommendation calls for investment in inter-disciplinary research, on social, legal and ethical
implications of AI that are relevant to public policy.

Due to the importance of data to the AI system lifecycle, a key element in ensuring further and better
AI R&D is the availability of open, accessible and representative datasets that do not compromise
privacy and personal and consumer data protection, intellectual property rights and other important
rights. In particular, while it may be impossible to achieve a completely “bias free” environment, by
providing (and providing incentives for) representative datasets that are publicly available,
governments can contribute to mitigating the risks of inappropriate bias in AI systems. For example,
AI systems using datasets that are not sufficiently representative in accordance with the system’s
intended use, even without the intention to discriminate.

This recommendation complements the recommendation on fostering a digital ecosystem for AI (2.2),
since long-term investment in digital technologies and infrastructure and mechanisms for sharing AI

OECD AI Principles | 8
knowledge are means to fostering this digital ecosystem. In particular, investment in open, accessible
and representative datasets facilitates sharing of AI knowledge.

Fostering an including AI-enabling ecosystem (Principle 2.2)

Governments should foster the development of, and access to, an inclusive, dynamic,
sustainable, and interoperable digital ecosystem for trustworthy AI. Such an ecosystem
includes inter alia, data, AI technologies, computational and connectivity infrastructure,
and mechanisms for sharing AI knowledge, as appropriate. In this regard, governments
should consider promoting mechanisms, such as data trusts, to support the safe, fair,
legal and ethical sharing of data.

Rationale for this principle

The development of trustworthy AI requires an enabling ecosystem. This recommendation calls on
governments – engaging with the private sector as appropriate – to work towards providing, or
promoting the provision of, the infrastructure and digital technologies for AI and the mechanisms for
AI knowledge sharing, taking into account their national frameworks.

The necessary digital technologies and infrastructure include access to affordable high-speed
broadband networks and services, computing power and data storage – as well as supporting data-
generating technologies such as the Internet-of-Things (IoT). For example, recent AI advances can
be attributed, in part, to the exponential increase in computational speed including with graphics-
processing unit resources. Appropriate mechanisms for sharing AI knowledge, including data, code,
algorithms, models, research, and know-how, are also required to understand and participate in the
AI system lifecycle. Such mechanisms must respect privacy, intellectual property and other rights.
Open source tools and high-quality training datasets for managing and using AI, which allow for the
diffusion of AI technology and crowdsourcing solutions to software bugs play a key role in AI
development.

When developing means of data sharing, such as data trusts or trusted third parties, governments
should pay attention to risks related to data access and sharing: risks to individuals (including
consumers), organisations and countries of sharing data may include confidentiality and privacy
breaches, risks to intellectual property rights, data protection, competition and commercial interests,
as well as potential national security and digital security risks. When it comes to the datasets
themselves, and in conjunction with recommendation 2.1, governments are invited to promote and
utilise datasets that are as inclusive, diverse and representative as possible.

The recommendations for policymakers draw special attention to policies for small and medium-sized
enterprises (SMEs): to facilitate SME access to data, AI technologies and relevant infrastructure (such
as connectivity, computing capacities and cloud platforms) in order to foster digital entrepreneurship,
competition and innovation through the adoption of AI.

OECD AI Principles | 9
Shaping and enabling interoperable governance and policy environment for AI
(Principle 2.3)
Governments should promote an agile policy environment that supports transitioning
from the research and development stage to the deployment and operation stage for
trustworthy AI systems. To this effect, they should consider using experimentation to
provide a controlled environment in which AI systems can be tested, and scaled-up, as
appropriate. They should also adopt outcome-based approaches that provide flexibility
in achieving governance objectives and co-operate within and across jurisdictions to
promote interoperable governance and policy environments, as appropriate.
Governments should review and adapt, as appropriate, their policy and regulatory
frameworks and assessment mechanisms as they apply to AI systems to encourage
innovation and competition for trustworthy AI.

Rationale for this principle

This recommendation focuses on the policy environment that enables AI innovation, i.e., institutional,
policy and legal framework. It thus complements recommendation 2.2 on the necessary physical and
technological infrastructure. As inrecommendation 2.2, governments are called to pay particular
attention to SMEs.

Considering the fast pace of AI developments, setting a policy environment that is flexible enough to
keep up with developments and promote innovation yet remains safe and provides legal certainty is
a significant challenge. This recommendation seeks to address this challenge by identifying means
for improving the adaptability, reactivity, versatility and enforcement of policy instruments, in order
to responsibly accelerate the transition from development to deployment and where relevant,
commercialisation. In fostering the use of AI, a human-centred approach should be taken.

The recommendation highlights the role of experimentation as a means to provide controlled and
transparent environments in which AI systems can be tested and in which AI-based business models
that could promote solutions to global challenges can flourish. Policy experiments can operate in
“start-up mode” whereby experiments are deployed, evaluated and modified, and then scaled up or
down, or abandoned, depending on the test outcomes.

Finally, the recommendation acknowledges the importance of oversight and assessment mechanisms
to complement policy frameworks and experimentation. In this respect, there may be room for
encouraging AI actors to develop self-regulatory mechanisms, such as codes of conduct, voluntary
standards and best practices. Together with the OECD Guidelines for Multinational Enterprises
(MNE)1, such initatives can help guide AI actors through the AI lifecycle, including for monitoring,
reporting, assessing and addressing harmful effects or misuse of AI systems. To the extent possible
and relevant, these mechanisms should be transparent and public.

1Available at: [Link]

OECD AI Principles | 10
Building human capacity and preparing for labour market transformation
(Principle 2.4)
Governments should work closely with stakeholders to prepare for the transformation of
the world of work and of society. They should empower people to effectively use and
interact with AI systems across the breadth of applications, including by equipping them
with the necessary skills.
Governments should take steps, including through social dialogue, to ensure a fair
transition for workers as AI is deployed, such as through training programmes along the
working life, support for those affected by displacement, including through social
protection, and access to new opportunities in the labour market.
Governments should also work closely with stakeholders to promote the responsible use
of AI at work, to enhance the safety of workers, the quality of jobs and of public services,
to foster entrepreneurship and productivity, and aim to ensure that the benefits from AI
are broadly and fairly shared.

Rationale for this principle

AI is broadly expected to change the nature of many aspects of life, as it diffuses across sectors. This
is particularly true in the context of labour, employment and the workplace – where AI will
complement humans in some tasks, replace them in others and generate new types of jobs and work
organisation. If poorly managed, these labour market transformations could have significant
economic and social costs. In managing these transitions and ensuring they are fair, policy makers –
together with stakeholders such as social partners, employer organisations and trade unions – will
need to consider questions around social protection, education programmes, skills development,
labour market regulation, public employment services, industrial policy, and taxation – as well as the
financing of transitions.

Managing fair transitions requires policies for life-long learning, skills development and training that
would allow people, and workers (in different contractual contexts) in particular, to interact with AI
systems, adapt to AI-generated changes and access new opportunities in the labour market. This
includes the skills required of AI practitioners (which are currently in shortage) and those needed for
other workers (such as doctors or lawyers) to be able to leverage AI in their areas of expertise, so that
AI augments human capabilities. In parallel, skills development policies will need to focus on the
distinctly human aspects necessary to complement AI systems, such as judgment, creative and critical
thinking and interpersonal communication.

Additionally, AI-led changes to the labour market may require adapting or adopting labour standards
and agreements between management and workers, where applicable, to reflect these changes,
address possible challenges to equality, diversity and fairness (posed, for example, by data collection
and processing), and to reinforce reliable, safe and productive workplaces. This can be achieved
through a combination of regulation, social dialogue and collective bargaining1. It is important to
allow for flexibility at the workplace while safeguarding workers’ autonomy and job quality.

OECD AI Principles | 11
1OECD (2019), OECD Employment Outlook 2019: The Future of Work, OECD Publishing,
Paris, [Link]

International co-operation for trustworthy AI (Principle 2.5)

Governments, including developing countries and with stakeholders, should actively co-
operate to advance these principles and to progress on responsible stewardship of
trustworthy AI.
Governments should work together in the OECD and other global and regional fora to
foster the sharing of AI knowledge, as appropriate. They should encourage international,
cross-sectoral and open multi-stakeholder initiatives to garner long-term expertise on
AI.
Governments should promote the development of multi-stakeholder, consensus-driven
global technical standards for interoperable and trustworthy AI.
Governments should also encourage the development, and their own use, of
internationally comparable indicators to measure AI research, development and
deployment, and gather the evidence base to assess progress in the implementation of
these principles.

Rationale for this principle

This recommendation calls for international co-operation among governments and with stakeholders,
to address the global opportunities and challenges of AI. Such co-operation includes advancing the
implementation and dissemination of these principles and policies across OECD and partner
countries, including developing countries and least developed countries, and with other stakeholders.

International co-operation can leverage international and regional fora to share AI knowledge in order
to build long-term expertise on AI; to develop technical standards for interoperable and trustworthy
AI; and to develop, disseminate and use metrics to assess the performance of AI systems, such as
accuracy, efficiency, advancement of societal goals, fairness and robustness. It can also contribute to
transborder flows of data with trust, that safeguard security, privacy, intellectual property, human
rights and democratic values and that are key to AI innovation.

OECD AI Principles | 12