must satisfy the organisational, quality management, resources and process requirements that are necessary to fulfil their tasks, as well as suitable cybersecurity requirements.
must ensure confidence in their performance, and in the results of the conformity assessment activities that they conduct.
must be independent of the provider of a high-risk AI system in relation to which they perform conformity assessment activities.
must be organised and operated so as to safeguard the independence, objectivity and impartiality of their activities.
must document and implement a structure and procedures to safeguard impartiality and to promote and apply the principles of impartiality throughout their organisation, personnel and assessment activities.
must have documented procedures in place ensuring that their personnel, committees, subsidiaries, subcontractors and any associated body or personnel of external bodies maintain the confidentiality of the information which comes into their possession during the performance of conformity assessment activities.
must observe professional secrecy with regard to all information obtained in carrying out their tasks under this Regulation.
must inform the other notified bodies of quality management system approvals which they have refused, suspended or withdrawn.
must inform the other notified bodies of Union technical documentation assessment certificates or any supplements thereto which they have refused, withdrawn, suspended or otherwise restricted.
must provide the other notified bodies carrying out similar conformity assessment activities covering the same types of AI systems with relevant information on issues relating to negative and, on request, positive conformity assessment results.
must safeguard the confidentiality of the information that they obtain.
must examine the technical documentation.
must grant full access to the training, validation, and testing data sets used, where relevant.
must require the provider to supply further evidence or carry out further tests, if necessary.
must carry out adequate tests, if necessary.
must grant access to the training and trained models of the AI system, including its relevant parameters, upon a reasoned request.
must notify the provider or its authorised representative of the decision.
Get notified when new obligations, deadlines, or guidance are added — or ask us how the caveauAI Compliance Workbench can track this for your organization.