INTERNATIONAL REGULATORY BODY FOR AI

International Regulatory Body for Artificial Intelligence

From ethics to safety to accountability, we cover all aspects of AI governance. Set up your IRBAI account to manage registrations, applications, certifications and audit verifications

Open account

ai governance

IRBAI has identified over 1,489 high-risk AI deployments across 110+ application domains and 10 critical sectors

Our organization conducts research, develops guidelines and regulations, and provides certification services to ensure that AI systems meet our standards for safety, ethics, transparency, and compliance. Our organization is dedicated to ensuring that the benefits of AI are realized while also mitigating its potential risks.

★★★★★

SAFE

Our regulatory body works to ensure that AI is developed and used in a way that prioritizes the safety of individuals and society as a whole. We establish safety standards to prevent harm caused by AI systems.

★★★★★

ETHICAL

We believe that AI development and use should be guided by ethical principles that prioritize fairness, accountability. Our organization established ethical guidelines that AI developers and users must adhere to.

★★★★★

TRANSPARENT

We ensure that AI systems are built in a way that allows for the explanation of their decisions and actions. We encourage developers and users to be open and transparent about their use of AI.

★★★★★

COMPLIANT

We establish and enforce regulations to ensure that AI development and use is compliant with ethical requirements. Our organization works closely with industry experts, policymakers and enforcement agencies.

Call to action

Join us to promote the responsible use of AI

You will be part of a global network of experts working to shape the future of AI and its impact on society. We are looking forward to working with AI developers, auditors, consultants and trainers.

Become IRBAI Member

assessment process

How we assess risks before granting IRBAI certifications?

At IRBAI, we grant different certifications to organizations and AI systems based on a thorough evaluation of their potential risks and impacts. Organizations without approprieate certifications cannot provide AI services nor products to the public.

Our assessment criteria include the following factors

Human Involvement & Oversight

The level of human involvement and oversight in the development, deployment, and operation of the AI system or device. This includes the level of transparency and explainability of the AI system.

Cybersecurity & Privacy Risks

The potential for the AI system or device to be vulnerable to cyberattacks, data breaches, and other security risks. This includes assessing unauthorized and authorised access to sensitive data and information and detection of infringement.

Level of Disruption

This criterion evaluates the number of people who may be impacted. For example, an AI system used in healthcare may have a direct impact on a single patient, while an AI system used in finance may impact companies

Financial Harm

The potential for the AI to cause financial harm to individuals or organizations, including the potential for fraud, manipulation, and market instability, money laundering or other illegal activities.

Mental Harm

The potential for the AI system or device to cause mental harm to humans, including psychological distress, anxiety, trauma, loss of autonomy, ability to make life decisions and loss of privacy.

Physical Harm

The potential for the AI system or device to cause physical harm to humans, including injury or death. This includes assessing the safety and reliability of the AI system or device, as well as its potential impact on public safety.

Environmental Impact

Impact of the AI system or device on the environment, including energy consumption, waste generation, and carbon footprint. This includes assessing the system’s energy efficiency and sustainability.

International Security

Implications of the AI system or device on international relations, including issues related to national security, diplomacy, global governance and cross-border data flow.

regulatory framework

How We Regulate AI Models

Regulatory risk does not follow from a model’s headline capability. IRBAI assesses each intersection of a model’s form and behaviour, and certifies those that carry risk.

01 · Risk at intersections

A given model form may present low risk in one behaviour and significant risk in another. Assessment is conducted at the level of the individual intersection, not the broad category.

02 · Capability ≠ risk

A narrow system of high capability may present limited risk; a less capable but general and autonomous system may present greater risk. Risk is assessed as a function of generality, autonomy, and access.

03 · Layered certification

A deployed model may require multiple certificates, one for each risk intersection it occupies. Certification is applied in layers corresponding to the structure of the system.

Explore the framework

INDUSTRY COMPLIANCE