Skip to content

Compliance, governance & resilience

Reviewing AI use and establishing accountability

An inventory of AI systems, risk classification and internal rules — taking into account the AI Act requirements that apply to the concrete use case.

IVSR helps the organization establish where and for what purpose it uses artificial intelligence, what role it plays in each system, and which legal, security and ethical risks it must manage.

Particular attention goes to AI in security, investigations, HR procedures, analytics and decision support. The work takes into account the requirements of the AI Act, the GDPR and sector rules that apply to the concrete system and use.

Possible scope

The scope follows the questions the engagement must answer — not the other way round.

  1. 01An inventory of AI systems and models, providers, data and purposes of use.
  2. 02Initial risk classification and determination of the organization's role for each system.
  3. 03Review of prohibited and especially sensitive uses.
  4. 04Rules for permitted generative-AI use and the protection of confidential data.
  5. 05Assessment of data quality, bias, explainability, traceability and human oversight.
  6. 06A procedure for approving new uses, recording incidents and periodic review.
  7. 07Training leadership and staff in safe and accountable use.

You receive

An AI-system register, a classification of uses, internal rules, a risk assessment, a procurement checklist, a training plan and a measures register.

Who it is for

For all organizations, especially those using AI in hiring, assessing people, security, biometrics, critical processes or the processing of sensitive data.

Limitations and boundaries

  • Obligations under the AI Act depend on the organization's role, the system's purpose and the timeline of individual provisions; blanket compliance promises are therefore not possible.
  • An AI system's output is not treated as standalone evidence or as a final decision about an individual.

Who performs what

IVSR
Research, advisory, methods and training
Next Sight
Technology platforms and technical analysis
The full role separation
Professional leadZvonimir Cvetko DamnjanovićDirector

Related services

Next step

A confidential introductory conversation

Every engagement starts with a confidential conversation in which we jointly verify the purpose, the authority and the feasibility. Only then do we propose the scope and the way of working.

Arrange a conversation