Inferensys

Glossary

Notified Body

An independent, accredited organization designated by an EU member state to conduct third-party conformity assessments of high-risk AI systems before they receive CE marking.
Risk analyst performing AI risk assessment on laptop, risk matrices visible, casual office risk session.
CONFORMITY ASSESSMENT

What is a Notified Body?

A Notified Body is an independent, accredited organization designated by an EU member state to conduct mandatory third-party conformity assessments of high-risk AI systems before they can receive CE marking and enter the European market.

A Notified Body is a third-party conformity assessment organization officially designated by an EU member state to evaluate whether high-risk AI systems comply with the essential requirements of the EU AI Act. These bodies possess the technical competence and independence to audit a provider's risk management system, technical documentation, and quality management system, ensuring the algorithm meets harmonized standards for safety and fundamental rights before market placement.

The specific tasks a Notified Body can perform are strictly limited by its accreditation scope, which defines the categories of AI systems it is authorized to assess. If a high-risk system undergoes a substantial modification, the provider must return to the Notified Body for a new conformity assessment. The body's decision is the gatekeeper for affixing the CE marking, serving as the definitive regulatory validation that the AI system is fit for the European market.

INDEPENDENT ASSESSMENT

Core Characteristics of a Notified Body

A notified body is a third-party conformity assessment organization designated by an EU member state to evaluate whether high-risk AI systems meet the essential requirements of the AI Act before they can receive CE marking and enter the European market.

01

Legal Designation and Competence

A notified body is not a self-appointed consultant but a legally designated entity. An EU member state's notifying authority formally assesses and designates the organization, then notifies the European Commission and other member states of its status.

  • Must demonstrate technical competence and impartiality for specific AI categories
  • Listed in the NANDO (New Approach Notified and Designated Organisations) database
  • Operates under the legal jurisdiction of its designating member state
02

Independence and Impartiality

The cornerstone of a notified body's credibility is its structural separation from the AI provider. It must be a third party free from any influence that could affect its technical judgment.

  • Cannot be the designer, manufacturer, or deployer of the AI system it assesses
  • Personnel must not engage in activities conflicting with their assessment objectivity
  • Financial independence from the commercial success of the assessed product is mandatory
03

Conformity Assessment Execution

The notified body performs the mandatory third-party assessment for high-risk AI systems, verifying compliance with the EU AI Act's essential requirements. This is the gatekeeping function before CE marking.

  • Audits the provider's quality management system and technical documentation
  • Evaluates the risk management system for completeness and effectiveness
  • May conduct product testing or require specific performance demonstrations
04

Accreditation Scope and Boundaries

A notified body is not authorized to assess all AI systems. Its accreditation scope strictly defines the categories of high-risk AI products it is competent to evaluate, based on its demonstrated technical expertise and resources.

  • Scope is formally defined in the designation notification
  • Covers specific AI technologies (e.g., biometrics, medical devices, safety components)
  • Operating outside the defined scope invalidates the conformity assessment
05

Ongoing Surveillance Obligations

The notified body's responsibility extends beyond the initial certification. It must conduct periodic audits and surveillance to ensure the AI system continues to comply with regulatory requirements throughout its lifecycle.

  • Verifies the provider's post-market monitoring system is functioning
  • Assesses substantial modifications to determine if re-certification is needed
  • Can suspend or withdraw certificates if non-compliance is identified
06

Liability and Competent Personnel

The notified body must maintain professional indemnity insurance and employ personnel with the specialized knowledge to evaluate complex AI systems. Assessors must understand both the underlying technology and the applicable legal framework.

  • Staff must possess deep knowledge of machine learning, data governance, and bias detection
  • Must understand the intersection of AI functionality and fundamental rights
  • Required to maintain confidentiality of all proprietary information obtained during assessment
NOTIFIED BODY ESSENTIALS

Frequently Asked Questions

Clear answers to the most common questions about the role, responsibilities, and operational mechanics of Notified Bodies under the EU AI Act.

A Notified Body is an independent, third-party organization officially designated by an EU member state to conduct conformity assessments of high-risk AI systems before they can be placed on the market. These bodies are the gatekeepers of the CE marking process, verifying that a provider's technical documentation, quality management system, and risk mitigation strategies meet the essential requirements of the AI Act. They are not government agencies but accredited private entities with demonstrated technical competence, impartiality, and the capacity to audit complex algorithmic systems. Their assessment results in a certificate that is a prerequisite for affixing the CE mark and legally commercializing the AI system within the European Union.

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.