By role · Deployer / user organisation
AI regulation for deployers and user organisations
A deployer uses an AI system under its own authority in the course of its business. Deployer duties are lighter than provider duties but they are the ones most organisations actually have: follow the instructions, keep humans in a position to oversee, retain logs, tell people when AI is used in decisions about them, and in some jurisdictions assess the impact before use.
- Jurisdictions
- 12
- Evidence items
- 75
What deployers get wrong
Treating a purchased system as the vendor's problem. The vendor's documentation is an input to the deployer's own oversight, notice and logging duties, not a substitute for them, and a deployer who materially changes a system can become its provider.
The short list
An inventory of the systems in use, a named owner for each, the vendor documentation on file, a human oversight arrangement that can actually intervene, notices to affected people, and logs kept for the period the instrument names. The controls below cover that list and say which duty each one serves.
Which controls meet these duties?
Sorted by how many of the duties on this page each control satisfies, so the ones worth building first are at the top. A control page lists every other duty it serves, in every jurisdiction.
| Control | Satisfies | Supports | Owner · frequency |
|---|---|---|---|
| AI interaction and use disclosure notices Process | 14 | 3 | Product owner · at launch and on material change |
| Privacy and data-protection controls for AI Process | 8 | 4 | Data protection officer · once per ai system |
| Decision explanation, human review and appeal route Process | 7 | 3 | Customer operations lead · once per ai system |
| AI governance policy and accountability structure Policy | 7 | 2 | Executive sponsor for AI · annual |
| AI risk assessment and lifecycle risk register Process | 5 | 5 | AI system owner · once per ai system |
| AI incident management and regulatory reporting Process | 5 | 1 | Incident coordinator · continuous |
| AI impact and fundamental-rights impact assessment Process | 4 | 6 | AI system owner · once per ai system |
| Prohibited and unacceptable-use screening gate Process | 4 | 1 | AI governance lead · once per ai system |
| Synthetic content labelling and provenance marking Technical measure | 4 | 0 | Engineering lead · continuous |
| Human oversight design and override procedure Process | 3 | 7 | AI system owner · once per ai system |
| Accuracy, robustness, fairness and security testing Technical measure | 3 | 5 | Quality or testing lead · at launch and on material change |
| Data governance and dataset documentation Process | 2 | 4 | Data governance lead · once per ai system |
| Post-deployment monitoring and drift detection Technical measure | 2 | 3 | AI system owner · continuous |
| Technical documentation, model cards and instructions for use Process | 1 | 3 | Product or model owner · at launch and on material change |
| Automatic event logging and record retention Technical measure | 1 | 3 | Engineering lead · continuous |
| AI literacy and role-based training programme Training programme | 1 | 3 | Learning and development lead · annual |
| Contractual allocation of AI duties across the supply chain Contractual term | 1 | 1 | Legal counsel · once per ai system |
| Public-sector AI use-case register and algorithmic transparency Process | 1 | 1 | Agency AI officer · annual |
| Conformity assessment, declaration and registration Process | 1 | 0 | Regulatory compliance lead · once per ai system |
| Model release and change-management gate Process | 1 | 0 | Release manager · at launch and on material change |
| AI system inventory and classification Process | 0 | 9 | AI governance lead · continuous |
| Vendor and third-party AI due diligence Process | 0 | 4 | Procurement or vendor risk lead · once per ai system |
| Adversarial and red-team testing for generative AI Technical measure | 0 | 2 | AI security or safety lead · at launch and on material change |
| Training-data provenance and copyright register Process | 0 | 1 | Model development lead · at launch and on material change |
| Quality management system for AI development and supply Policy | 0 | 1 | Quality lead · annual |
Which duties are recorded?
Every published duty whose record names this audience. It is the recorded set, not every rule in the world; a jurisdiction missing here may simply not be mapped yet (open gaps).
Australia 3 duties
-
VoluntaryAustralian Voluntary AI Safety Standard · Guardrails 1 and 2Establish accountability processes and a risk-management process (guardrails 1 and 2)
-
VoluntaryAustralian Voluntary AI Safety Standard · Guardrails 7, 8 and 9Provide contestability, supply-chain transparency and records (guardrails 7 to 9)
-
VoluntaryAustralian Voluntary AI Safety Standard · Guardrails 4, 5 and 6Test and monitor systems, enable human control, and be transparent with users (guardrails 4 to 6)
Colorado (United States) 7 duties
-
Legal requirementColorado AI Act · C.R.S. 6-1-1704applies from 30 Jun 2026Deployers and developers must disclose to consumers that they are interacting with an AI system
-
Legal requirementColorado AI Act · C.R.S. 6-1-1703(3)applies from 30 Jun 2026Deployers must complete impact assessments for high-risk AI
-
Legal requirementColorado AI Act · C.R.S. 6-1-1703(2)applies from 30 Jun 2026Deployers must implement a risk management policy and programme
-
Legal requirementColorado AI Act · C.R.S. 6-1-1703(7)applies from 30 Jun 2026Deployers must notify the Attorney General of discovered algorithmic discrimination
-
Legal requirementColorado AI Act · C.R.S. 6-1-1703(5)applies from 30 Jun 2026Deployers must publish a statement about the high-risk AI systems they use
-
Legal requirementColorado AI Act · C.R.S. 6-1-1703(1)applies from 30 Jun 2026Deployers must use reasonable care to avoid algorithmic discrimination
-
Legal requirementColorado AI Act · C.R.S. 6-1-1703(4)applies from 30 Jun 2026Notify consumers and explain adverse consequential decisions
European Union 20 duties
-
Legal requirementEU AI Act · Article 27applies from 2 Aug 2026Carry out a fundamental rights impact assessment before deployment
-
Legal requirementEU AI Act · Articles 43, 47, 48 and 49; Annex VIIIapplies from 2 Aug 2026Complete conformity assessment, CE marking and EU database registration
-
Legal requirementEU AI Act · Article 50(4)applies from 2 Aug 2026Deployers must disclose deepfakes and AI-generated text published on matters of public interest
-
Legal requirementEU AI Act · Article 26(4)applies from 2 Aug 2026Deployers must ensure input data they control is relevant and representative
-
Legal requirementEU AI Act · Article 86applies from 2 Aug 2026Deployers must explain individual decisions taken with high-risk AI on request
-
Legal requirementEU AI Act · Article 26(5)applies from 2 Aug 2026Deployers must monitor high-risk AI, suspend use on risk and report serious incidents
-
Legal requirementEU AI Act · Article 26(11)applies from 2 Aug 2026Deployers must tell natural persons that a high-risk AI system is used in decisions about them
-
Legal requirementEU AI Act · Article 26(9)applies from 2 Aug 2026Deployers must use the provider's transparency information in their data protection impact assessment
-
Legal requirementEU AI Act · Article 50(3)applies from 2 Aug 2026Deployers of emotion recognition or biometric categorisation must inform exposed persons
-
Legal requirementEU AI Act · Article 25(1) and 25(2)applies from 2 Aug 2026Deployers, distributors and importers must assume provider duties when they rebrand or substantially modify high-risk AI
-
Legal requirementEU AI Act · Article 12; Article 26(6) for deployersapplies from 2 Aug 2026Design high-risk systems to log events automatically
-
Legal requirementEU AI Act · Article 50applies from 2 Aug 2026Disclose AI interaction and label synthetic content
-
Legal requirementEU AI Act · Article 5applies from 2 Feb 2025Do not deploy or provide AI for prohibited practices
-
Legal requirementEU AI Act · Article 26(7)applies from 2 Aug 2026Employers must inform workers and their representatives before using high-risk AI at work
-
Legal requirementEU AI Act · Article 14; Article 26(2) for deployersapplies from 2 Aug 2026Enable and assign effective human oversight
-
Legal requirementEU AI Act · Article 4applies from 2 Feb 2025Ensure AI literacy of staff operating AI systems
-
Legal requirementEU AI Act · Article 26(10)applies from 2 Aug 2026Law-enforcement deployers must obtain authorisation for post-remote biometric identification and report annually
-
Legal requirementEU AI Act · Article 26(8); Article 49(3) and 49(4)applies from 2 Aug 2026Public authorities must register their use of high-risk AI and must not use unregistered systems
-
Legal requirementEU AI Act · Article 73applies from 2 Aug 2026Report serious incidents to market surveillance authorities
-
Legal requirementEU AI Act · Article 26applies from 2 Aug 2026Use high-risk AI as instructed, monitor it and inform affected people
India 4 duties
-
Legal requirementIndia DPDP Act · Section 8(5) and 8(6); DPDP Rules on breach intimationImplement reasonable security safeguards and notify breaches
-
Legal requirementIndia DPDP Act · Sections 4 to 7Process personal data only with valid consent or a legitimate use, after notice
-
Legal requirementIndia DPDP Act · Section 10Significant Data Fiduciaries must appoint a DPO and run impact assessments and audits
-
VoluntaryIndia AI Governance Guidelines · Guiding principles and recommendations sectionsAdopt the guiding principles and risk-based governance (voluntary)
Nepal 1 duty
-
Legal requirementNepal Privacy Act 2075 · Chapter on collection and protection of personal information (reviewer to cite sections)Collect and use personal information only with consent and for the stated purpose
New York (United States) 5 duties
-
Legal requirementNYC Local Law 144 (automated employment decision tools) · NYC Administrative Code Section 20-871(b)(3); 6 RCNY Section 5-303applies from 5 Jul 2023Employers and employment agencies must disclose the data collected and their retention policy for the tool
-
Legal requirementNYC Local Law 144 (automated employment decision tools) · NYC Administrative Code Section 20-871(b)(1); 6 RCNY Section 5-303applies from 5 Jul 2023Employers and employment agencies must let candidates request an alternative selection process or accommodation
-
Legal requirementNYC Local Law 144 (automated employment decision tools) · NYC Administrative Code Section 20-871(b)(1) and (b)(2); 6 RCNY Section 5-303applies from 5 Jul 2023Employers and employment agencies must notify candidates and employees before an automated tool is used
-
Legal requirementNYC Local Law 144 (automated employment decision tools) · NYC Administrative Code Section 20-871(a)(1); 6 RCNY Section 5-301applies from 5 Jul 2023Employers and employment agencies must obtain an independent bias audit before using an automated employment decision tool
-
Legal requirementNYC Local Law 144 (automated employment decision tools) · NYC Administrative Code Section 20-871(a)(2); 6 RCNY Section 5-302applies from 5 Jul 2023Employers and employment agencies must publish a summary of the bias audit results
Singapore 6 duties
-
Legal requirementPDPC AI advisory guidelines · Advisory guidelines, sections on consent, business improvement and research exceptionsIdentify consent or an applicable PDPA exception before using personal data in AI
-
Legal requirementPDPC AI advisory guidelines · Advisory guidelines, section on notification obligationNotify individuals about the use of personal data in AI recommendations and decisions
-
VoluntarySingapore Model AI Governance Framework · Second edition, Part on human involvement in AI-augmented decision-makingDetermine the appropriate level of human involvement in AI decisions
-
VoluntarySingapore Model AI Governance Framework · Second edition, Part on internal governance structures and measuresEstablish internal governance structures and measures for AI
-
VoluntarySingapore Model AI Governance Framework · Second edition, Part on operations managementManage data quality, model development and monitoring across the lifecycle
-
VoluntarySingapore Model AI Governance Framework · Generative AI framework, dimensions on incident reporting and content provenanceReport incidents and mark AI-generated content (generative AI framework)
South Korea 7 duties
-
Legal requirementFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust · Article 31(2) and 31(3)applies from 22 Jan 2026AI business operators must label generative AI output and clearly flag realistic synthetic media
-
Legal requirementFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust · Article 31(1)applies from 22 Jan 2026AI business operators must notify users in advance that a product or service runs on high-impact or generative AI
-
Legal requirementFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust · Article 34(1)applies from 22 Jan 2026Operators of high-impact AI must be able to explain outputs and the main criteria behind them
-
Legal requirementFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust · Article 34(1)applies from 22 Jan 2026Operators of high-impact AI must ensure human management and supervision
-
Legal requirementFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust · Article 34(1)applies from 22 Jan 2026Operators of high-impact AI must establish and operate a risk management plan
-
Legal requirementFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust · Article 34(1)applies from 22 Jan 2026Operators of high-impact AI must prepare user-protection measures and keep records of their safety and trust measures
-
VoluntaryFramework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust · Article 35applies from 22 Jan 2026Operators of high-impact AI should assess its impact on fundamental rights before use
Texas (United States) 5 duties
-
Legal requirementTexas Responsible AI Governance Act (TRAIGA) · Business and Commerce Code Section 551.052applies from 1 Jan 2026Developers and deployers must not use AI to incite self-harm, harm to others or crime
-
Legal requirementTexas Responsible AI Governance Act (TRAIGA) · Business and Commerce Code Section 551.056applies from 1 Jan 2026Developers and deployers must not use AI with the intent to unlawfully discriminate against a protected class
-
Legal requirementTexas Responsible AI Governance Act (TRAIGA) · Business and Commerce Code Section 551.057applies from 1 Jan 2026Developers and distributors must not build AI intended to produce child sexual abuse material or unlawful sexual deepfakes
-
Legal requirementTexas Responsible AI Governance Act (TRAIGA) · Business and Commerce Code Section 551.051applies from 1 Jan 2026Government agencies must disclose to consumers that they are interacting with an AI system
-
Legal requirementTexas Responsible AI Governance Act (TRAIGA) · Business and Commerce Code Section 551.051applies from 1 Jan 2026Health-care providers must disclose the use of AI in patient services
United Arab Emirates 2 duties
-
Legal requirementUAE PDPL · Article on data protection impact assessment (reviewer to cite article number)Conduct a data protection impact assessment for high-risk processing using new technologies
-
Legal requirementUAE PDPL · Article on data-subject rights relating to automated processing (reviewer to cite article number)Respect the right to object to automated decision-making without human intervention
United Kingdom 7 duties
-
Legal requirementICO AI guidance · UK GDPR Article 22 as amended by the Data (Use and Access) Act 2025Apply safeguards to solely automated decisions with significant effects
-
Legal requirementICO AI guidance · UK GDPR Article 35; ICO guidance, accountability and governance sectionCarry out a data protection impact assessment for high-risk AI processing
-
VoluntaryUK AI regulation framework · Principle 1, Part 3Ensure AI systems are safe, secure and robust throughout their lifecycle
-
VoluntaryUK AI regulation framework · Principle 4, Part 3Establish accountability and governance for AI
-
VoluntaryUK AI regulation framework · Principle 2, Part 3Provide appropriate transparency and explainability
-
VoluntaryUK AI regulation framework · Principle 5, Part 3Provide routes to contest AI outcomes and seek redress
-
VoluntaryUK AI regulation framework · Principle 3, Part 3Use AI in ways that are fair and do not discriminate unlawfully
United States 3 duties
-
VoluntaryNIST AI RMF · GOVERN functionEstablish AI governance policies, roles and accountability (Govern)
-
VoluntaryNIST AI RMF · MAP functionMap context, intended use and potential impacts (Map)
-
VoluntaryNIST AI RMF · MANAGE functionPrioritise, respond to and monitor AI risks (Manage)
What evidence would a reviewer expect?
- AI customer or deployer clause set Contract clause or supplier term
- AI data-flow and legal-basis record Register entry
- AI decision challenge and human review procedure Procedure or standard operating process
- AI governance forum minutes Governance meeting record
- AI impact assessment Impact assessment
- AI incident record Incident record
- AI incident response playbook Procedure or standard operating process
- AI intake and classification procedure Procedure or standard operating process
- AI interaction or use notice Disclosure or notice
- AI policy Policy document
- AI quality management system manual Policy document
- AI responsibility map Register entry
- AI supplier and component register Register entry
- AI supplier clause set Contract clause or supplier term
- AI supplier due-diligence assessment Supplier assessment
- AI system event logs Access or activity log
- AI system register Register entry
- AI system risk assessment Risk assessment
- AI training completion records Training record
- AI training curriculum and materials Policy document
- Adversarial findings tracker Risk register
- Adverse-decision explanation template Disclosure or notice
- Algorithmic transparency statement for one use case Disclosure or notice
- Board or executive approval of the AI policy Approval or sign-off record
- Challenge and reversal log Monitoring record
- Conformity evidence pack Technical documentation file
- Content labelling and provenance standard Procedure or standard operating process
- Contract clause index against the AI register Register entry
- Copyright and rights-reservation policy Policy document
- Data protection impact assessment for an AI system Data protection impact assessment
- Data quality and bias check report Evaluation or test report
- Dataset approval for use Approval or sign-off record
- Dataset documentation sheet Dataset documentation
- Declaration of conformity or certificate Conformity declaration or certificate
- Human oversight and override procedure Procedure or standard operating process
- Human-involvement design rationale Approval or sign-off record
- Impact assessment approval Approval or sign-off record
- Impact assessment procedure and template Procedure or standard operating process
- Incident report to an authority Regulatory filing or notification
- Independent data audit or DPO review Audit or assurance report
- Instructions for use Disclosure or notice
- Internal audit of the AI management system Audit or assurance report
- Inventory review and publication sign-off Approval or sign-off record
- Log integrity and retention check Audit or assurance report
- Log schema and retention standard Procedure or standard operating process
- Management review minutes Governance meeting record
- Model card or deployer information pack Model documentation
- Monitoring dashboard or periodic monitoring report Monitoring record
- Monitoring review decision Approval or sign-off record
- Notice catalogue Register entry
- Notice wording approval Approval or sign-off record
- Overseer training completion Training record
- Per-system AI risk register Risk register
- Post-market monitoring plan Procedure or standard operating process
- Pre-release test report Evaluation or test report
- Privacy notice section on AI use Disclosure or notice
- Prohibited-use screening record Approval or sign-off record
- Public AI use-case inventory Register entry
- Public summary of training content Disclosure or notice
- Red-team exercise report Evaluation or test report
- Red-team rules of engagement and scenario library Procedure or standard operating process
- Registration record in the relevant database Regulatory filing or notification
- Release and change-classification procedure Procedure or standard operating process
- Release or change approval record Approval or sign-off record
- Release test sign-off Approval or sign-off record
- Residual-risk acceptance Approval or sign-off record
- Risk-tier classification sign-off Approval or sign-off record
- Role-to-curriculum training matrix Procedure or standard operating process
- Screening list and escalation procedure Procedure or standard operating process
- Supplier onboarding decision Approval or sign-off record
- Technical documentation file Technical documentation file
- Test plan and acceptance criteria Procedure or standard operating process
- Training source register Register entry
- Visible AI-generated content label Disclosure or notice
- Watermark and provenance robustness test Evaluation or test report
Latest changes to these instruments
Texas TRAIGA takes effect
European Commission proposes Digital Omnibus adjustments to AI Act timelines
India notifies the Digital Personal Data Protection Rules, 2025
MeitY releases the India AI Governance Guidelines
Informational only, not legal advice. Verify every claim against the linked official sources and consult qualified counsel before acting.
Frequently asked questions
- Do deployer duties apply to internal tools?
- Usually yes when the tool makes or supports decisions about people, such as hiring, credit or access to services. Purely personal, non-professional use is typically excluded. The duty pages cite the scope article.
- What evidence should a deployer keep?
- The system register entry, the vendor's instructions and documentation, the oversight arrangement, the notices shown to people, the logs, and any impact assessment the instrument requires. The controls below list each item.