Artificial Intelligence Policy Tracker

AI risk management: AI obligations

Practical requirements extracted from policy instruments, with the source article, the actors they bind, evidence examples and original framework mappings. Legal requirements are marked; everything else is voluntary guidance.

Results

10 of 129 obligations match · Download these as CSV

At a glance · By category, for the records that match

Binding 8 legally binding 2 voluntary or guidance

Legal requirement AI risk management United States

Apply minimum risk-management practices to high-impact AI

OMB M-25-21 · Section 4

For AI whose output serves as a principal basis for decisions with significant effects on rights, safety or access to services, agencies must complete pre-deployment testing, an AI impact assessment, ongoing monitoring, operator training, human oversight and a mechanism for affected people to seek remedy, or stop using the AI.

Pending review (a factual check against the official source, not a legal review)
Legal requirement AI risk management Colorado (United States)

Deployers must implement a risk management policy and programme

Colorado AI Act · C.R.S. 6-1-1703(2)

Deployers of high-risk AI must implement a risk-management policy and programme governing deployment, specifying principles, processes and personnel used to identify, document and mitigate known or reasonably foreseeable risks of algorithmic discrimination, and reasonable in light of recognised frameworks such as the NIST AI RMF or ISO/IEC 42001.

Pending review (a factual check against the official source, not a legal review)
Legal requirement AI risk management Colorado (United States)

Deployers must use reasonable care to avoid algorithmic discrimination

Colorado AI Act · C.R.S. 6-1-1703(1)

A deployer of a high-risk AI system must use reasonable care to protect consumers from known or reasonably foreseeable risks of algorithmic discrimination. A deployer that runs the required risk-management programme, completes impact assessments, gives consumer notices and makes the required disclosures benefits from a rebuttable presumption of reasonable care; some deployers with fewer than 50 employees are relieved of parts of the programme and assessment duties when they rely on the developer's impact assessment.

Pending review (a factual check against the official source, not a legal review)
Legal requirement AI risk management Colorado (United States)

Developers must use reasonable care to avoid algorithmic discrimination

Colorado AI Act · C.R.S. 6-1-1702(1)

A developer of a high-risk AI system must use reasonable care to protect consumers from any known or reasonably foreseeable risk of algorithmic discrimination arising from the intended and contracted uses of the system. A developer that meets the documentation, public-statement and Attorney General disclosure duties in the rest of section 6-1-1702 benefits from a rebuttable presumption that it used reasonable care.

Pending review (a factual check against the official source, not a legal review)
Legal requirement AI risk management European Union

Establish a risk management system for high-risk AI

EU AI Act · Article 9

Providers of high-risk AI systems must establish, implement, document and maintain a continuous, iterative risk-management system across the system's lifecycle: identifying known and reasonably foreseeable risks to health, safety and fundamental rights, estimating and evaluating risks including from reasonably foreseeable misuse, evaluating post-market data, and adopting targeted risk-management measures, with testing before placing on the market.

Pending review (a factual check against the official source, not a legal review) Applies from 2 Dec 2027
Legal requirement AI risk management Washington State (United States)

Keep and run a protocol to detect suicidal ideation and self-harm and refer users to crisis resources

Washington AI companion chatbot law (ESHB 2225) · ESHB 2225 (Chapter 168, Laws of 2026), sec. 5(1) and (2)

An operator may not make available or deploy an AI companion chatbot unless it maintains and implements a protocol for detecting and addressing users' suicidal ideation or expressions of self-harm. The protocol must include reasonable methods to identify expressions of suicidal ideation or self-harm, eating disorders included; give automated or human-mediated responses that refer users to appropriate crisis resources such as a suicide hotline or crisis text line; and include reasonable measures to prevent the chatbot generating content that encourages self-harm or describes how to do it.

Pending review (a factual check against the official source, not a legal review) Applies from 1 Jan 2027
Legal requirement AI risk management New York (United States)

Offer an AI companion only if it contains a protocol to detect suicidal ideation or self-harm and refer users to crisis services

New York AI companion safeguards · N.Y. Gen. Bus. Law § 1701

It is unlawful for an operator to operate or provide an AI companion to a user unless the AI companion contains a protocol for taking reasonable efforts to detect and address suicidal ideation or expressions of self-harm by the user. The protocol must at least detect such expressions and, when it does, notify the user with a referral to crisis service providers, such as the 9-8-8 suicide prevention and behavioral health crisis hotline, a crisis text line or other appropriate crisis services. The Attorney General may seek an injunction and civil penalties of up to 15,000 dollars per day for a violation (section 1703).

Pending review (a factual check against the official source, not a legal review)
Legal requirement AI risk management South Korea

Operators of high-impact AI must establish and operate a risk management plan

AI Basic Act · Article 34(1)

An AI business operator that provides high-impact AI, or a product or service using it, must establish and operate a plan for managing the risks of that AI. High-impact AI is AI used in areas such as energy, drinking water, health care and medical devices, nuclear safety, biometric analysis for criminal investigation, decisions on recruitment or loans, transport, public services and education, where it may materially affect life, safety or fundamental rights.

Verified (a factual check against the official source, not a legal review) Applies from 22 Jan 2026
Voluntary guidance AI risk management United Kingdom

Ensure AI systems are safe, secure and robust throughout their lifecycle

UK AI regulation framework · Principle 1, Part 3

Regulators are asked to ensure AI systems function in a robust, secure and safe way, with risks continually identified, assessed and managed. In practice this is enforced through existing safety, security and data-protection law rather than a new duty.

Pending review (a factual check against the official source, not a legal review)
Voluntary guidance AI risk management United States

Prioritise, respond to and monitor AI risks (Manage)

NIST AI RMF · MANAGE function

Manage allocates resources to mapped and measured risks, plans responses including decommissioning, manages third-party risks, and documents post-deployment monitoring, incident response and communication.

Pending review (a factual check against the official source, not a legal review)

Frequently asked questions

What is an obligation on this site?

A single practical requirement pulled out of an instrument and stated on its own: keep a risk management system, log incidents, document training data, provide human oversight, and so on. Each one cites the article or section it comes from so you can check it against the source.

Does a voluntary obligation have legal force?

No, and every obligation is labelled either a legal requirement or voluntary guidance. Voluntary items still matter in practice, because procurement questionnaires and auditors ask about them, but only the binding ones carry legal consequence.

How do I find the obligations that apply to my organisation?

Filter by jurisdiction, category, actor, sector or use case. The applicability check asks a short set of questions and returns the duties that may reach you. It is an educational screen, not a legal determination, and it says so.

Why do some instruments have no obligations listed?

Because nobody has broken them out yet. Most instruments are recorded at summary level first; obligations are added jurisdiction by jurisdiction. The coverage and open-gaps pages publish exactly what is missing rather than hiding it.