Article 22: Authorized Representatives of Providers of High-Risk AI Systems | EU AI Act

Contributors

Anas Baig

Product Marketing Manager at Securiti

Semra Islam

Sr. Data Privacy Analyst

CIPM, CIPP/Europe

Article 22 of the AI Act provides information about the authorized representatives of providers of high-risk AI systems, including their duties and responsibilities.

Providers of high-risk AI systems based out of third-world countries must appoint an authorized representative by written mandates before making their systems available on the EU market. Such a representative must be empowered to perform the tasks as instructed by the provider.

The authorized representative must perform all such tasks mandated by the provider. It shall provide a copy of the mandate to the market surveillance authorities upon request in one of the official languages of the EU as indicated by the competent authority.

The mandate shall empower the authorized representative to be addressed by the competent authorities, in addition to or instead of the provider, on all issues concerning compliance with the AI Act.

The mandate will empower the authorized representative to perform any of the following tasks:

  • Verify that the technical documentation and the EU declaration of conformity are drawn up per the requirements of the Act;
  • Ensure the appropriate conformity assessment procedure has been carried out by the provider;
  • Keep the following documents for a period of 10 years after the high-risk AI system is put into service or placed on the market, in case requested by the national competent authorities and national authorities or bodies that supervise the application of EU harmonization laws, or are referred in other EU laws that might be relevant for the high-risk AI systems:
    • Contact details of the provider;
    • Copy of the EU declaration of conformity;
    • Technical documentation;
    • The certificate issued by the notified body.
  • Provide a national competent authority with all the information and documentation necessary when requested to demonstrate the conformity of their high-risk AI system, including access to relevant logs generated by the high-risk AI system;
  • Cooperate with the competent authorities when requested in any actions being taken in relation to the high-risk AI system, especially if the actions aim to reduce or mitigate the risks posed by the high-risk AI system;
  • Comply with the registration requirements per Article 49, or if the registration is carried out by the provider itself, ensure their name, address and contact details are correct for registration purposes.

The authorized representative must terminate the mandate if they have reasons to believe that the provider is acting against the provisions of the AI Act and may be in violation of their obligations. Additionally, they must inform the relevant market surveillance authority and the relevant notified body about the termination of the mandate and the reasons behind the termination.

Join Our Newsletter

Get all the latest information, law updates and more delivered to your inbox


Share

More Stories that May Interest You
Videos
View More
Rehan Jalil, Veeam on Agent Commander : theCUBE + NYSE Wired: Cyber Security Leaders
Following Veeam’s acquisition of Securiti, the launch of Agent Commander marks an important step toward helping enterprises adopt AI agents with greater confidence. In...
View More
Mitigating OWASP Top 10 for LLM Applications 2025
Generative AI (GenAI) has transformed how enterprises operate, scale, and grow. There’s an AI application for every purpose, from increasing employee productivity to streamlining...
View More
Top 6 DSPM Use Cases
With the advent of Generative AI (GenAI), data has become more dynamic. New data is generated faster than ever, transmitted to various systems, applications,...
View More
Colorado Privacy Act (CPA)
What is the Colorado Privacy Act? The CPA is a comprehensive privacy law signed on July 7, 2021. It established new standards for personal...
View More
Securiti for Copilot in SaaS
Accelerate Copilot Adoption Securely & Confidently Organizations are eager to adopt Microsoft 365 Copilot for increased productivity and efficiency. However, security concerns like data...
View More
Top 10 Considerations for Safely Using Unstructured Data with GenAI
A staggering 90% of an organization's data is unstructured. This data is rapidly being used to fuel GenAI applications like chatbots and AI search....
View More
Gencore AI: Building Safe, Enterprise-grade AI Systems in Minutes
As enterprises adopt generative AI, data and AI teams face numerous hurdles: securely connecting unstructured and structured data sources, maintaining proper controls and governance,...
View More
Navigating CPRA: Key Insights for Businesses
What is CPRA? The California Privacy Rights Act (CPRA) is California's state legislation aimed at protecting residents' digital privacy. It became effective on January...
View More
Navigating the Shift: Transitioning to PCI DSS v4.0
What is PCI DSS? PCI DSS (Payment Card Industry Data Security Standard) is a set of security standards to ensure safe processing, storage, and...
View More
Securing Data+AI : Playbook for Trust, Risk, and Security Management (TRiSM)
AI's growing security risks have 48% of global CISOs alarmed. Join this keynote to learn about a practical playbook for enabling AI Trust, Risk,...

Spotlight Talks

Spotlight 59:11
Data Controls for AI: Findings from the 2026 GigaOm DSPM Research
Watch Now View
Spotlight 1:02:06
Consent by proxy: When AI agents start deciding for us
Watch Now View
Spotlight 1:00:41
Future-Proofing for the Privacy Professional
Watch Now View
Spotlight 50:52
From Data to Deployment: Safeguarding Enterprise AI with Security and Governance
Watch Now View
Spotlight 11:29
Not Hype — Dye & Durham’s Analytics Head Shows What AI at Work Really Looks Like
Not Hype — Dye & Durham’s Analytics Head Shows What AI at Work Really Looks Like
Watch Now View
Spotlight 11:18
Rewiring Real Estate Finance — How Walker & Dunlop Is Giving Its $135B Portfolio a Data-First Refresh
Watch Now View
Spotlight
Choosing the Right DSPM: An Industry Analyst’s Perspective
Watch Now View
Spotlight 13:38
Accelerating Miracles — How Sanofi is Embedding AI to Significantly Reduce Drug Development Timelines
Sanofi Thumbnail
Watch Now View
Spotlight 10:35
There’s Been a Material Shift in the Data Center of Gravity
Watch Now View
Spotlight 14:21
AI Governance Is Much More than Technology Risk Mitigation
AI Governance Is Much More than Technology Risk Mitigation
Watch Now View
Latest
Australia’s Office of AI: Why Annual Audits Miss What Your AI Can Reach View More
Australia’s Office of AI: Why Annual Audits Miss What Your AI Can Reach
Picture this: a fictional but entirely plausible scenario. An Australian financial institution's AI systems spend six months accessing a customer data repository nobody has...
View More
A Complete DSPM Needs Classification and Context
Classification is one of the core functions a DSPM program handles, and it usually runs in tandem with discovery, since together they form the...
The EU Cyber Resilience Act (CRA): A Complete Guide for Compliance View More
The EU Cyber Resilience Act (CRA): A Complete Guide for Compliance
Here’s what you need to know about the EU’s Cyber Resilience Act (CRA), including the best solutions to aid your compliance efforts. Read on...
What is Access Control? Definition, Types, & Components View More
What is Access Control? Definition, Types, & Components
Discover what access control is, how it works, types, components, importance in ensuring regulatory compliance, and much more.
Stop Storing Risk: An Executive's Guide to ROT Data Minimization View More
Stop Storing Risk: An Executive’s Guide to ROT Data Minimization
An executive's guide to reducing redundant, obsolete, and trivial (ROT) data to cut storage costs, shrink your attack surface, and improve compliance.
The Context Layer for Data+AI Security View More
The Context Layer for Data+AI Security
Discover how Securiti’s DataAI Command Graph connects data, identity, cloud, and AI findings to uncover contextual risk and toxic combinations.
The Toxic Combination Problem in DataAI Risks View More
The Toxic Combination Problem in DataAI Risks
Discover how siloed security alerts create hidden toxic risk combinations and how correlated context helps reduce alert fatigue and uncover compound risks faster.
The Cloud Storage Bill Nobody Reads View More
The Cloud Storage Bill Nobody Reads
Hidden cloud storage costs add up fast. Learn how redundant, obsolete, and trivial data drives unnecessary spend, expands risk, and why automated data minimization...
View More
Take the Data Risk Out of AI
Learn how to prepare enterprise data for safe Gemini Enterprise adoption with upstream governance, sensitive data discovery, and pre-index policy controls.
View More
Navigating HITRUST: A Guide to Certification
Securiti's eBook is a practical guide to HITRUST certification, covering everything from choosing i1 vs r2 and scope systems to managing CAPs & planning...
What's
New