Research
Innovation readiness, documentation, knowledge management, and continuous development
AAM-005
Trustworthy AI
Assessment
Can you place justified trust in your AI?
Trust in AI does not arise from performance, compliance, or good intentions alone. It requires reasons that can be understood and justified – across technology, organisation, and people.
The Trustworthy AI Assessment examines what this trust is actually based on. We bring together available Evidence, Known Limitations, Uncertainty, Governance, and effective Controls into a traceable Trust Profile. This makes visible which assumptions are sufficiently justified – and where further Evidence is needed.
Discuss the Trustworthy AI Assessment →01
Why
AI systems increasingly influence business processes, interactions, and decisions. The greater this influence becomes, the more important it is to understand whether a solution can be used not only effectively, but also securely, transparently, responsibly, and sustainably.
Individual technical tests or governance measures answer this question only in part. The Trustworthy AI Assessment connects different perspectives and makes visible what existing trust is actually based on – and where assumptions are not yet sufficiently supported by Evidence.
Trust is stronger when its basis is visible.
02
What
Five dimensions. One holistic view of trustworthiness.
People
Human-AI Interaction, transparency, competence, Human Oversight, and organisational integration
Quality
Data and model quality, robustness, reproducibility, monitoring, and improvement
Governance
Responsibilities, Human Oversight, policies, traceability, and risk acceptance
Security
AI Security, identities, data flows, access protection, and resilience
03
How
We define the context and scope, analyse available Evidence, and involve relevant stakeholders. We then make visible what trust is based on today, which Known Limitations and Uncertainties remain, and where additional Evidence is needed.
Scope
Define AI systems, context, objectives, and assessment scope
Evidence
Analyse documentation, architecture, processes, and existing results
Engage
Engage stakeholders and understand how the AI is used in practice
Assess
Assess Research, People, Quality, Governance, and Security holistically
Challenge
Discuss findings, challenge assumptions, and prioritise improvements
04
Knowledge
After the assessment, you have a much clearer understanding of the basis on which trust in your AI can be justified:
Which Evidence actually supports trust in the AI system under assessment
Which Known Limitations need to be considered
Which Uncertainties remain despite the available Evidence
Which assumptions are not yet sufficiently supported
Which additional Evidence is needed to close critical Trust Gaps
Typical deliverablesTrustworthy AI Assessment ReportTrust ProfileEvidence BaseKnown Limitations & UncertaintyEvidence & Trust GapsPrioritised Recommendations
05
scip AG
We have been researching Trust in AI since trust was still rarely part of the AI discussion.
scip combines more than twenty years of cybersecurity experience with our own research into AI, Trust, and Human-AI Interaction. As early as 2018, we examined Trust in AI from a systemic perspective and have continued to develop this research ever since.
Our assessment therefore does not rely solely on regulatory requirements or technical best practices. We combine Research, practical project experience, and interdisciplinary thinking to determine whether trust in a specific AI solution can be justified in a traceable way – and what that trust is actually based on.
2018+
5
Evidence
06
Perspective
AAM is not a linear process. Depending on the context, other questions may become relevant and different capabilities can be applied where they add value.
07
Insights
Trustworthy AI is not a new compliance term for scip. Our research into Trust and AI goes back several years and forms a central foundation of this module.
On Trust in AI – A Systemic Approach →ISO/IEC TR 24028:2020 — Overview of Trustworthiness in Artificial Intelligence →Trust and AI – Three Wrong Questions →ISO/IEC 42005:2025 — AI System Impact Assessment →Eight Questions on the Path to Trustworthy AI – Trust Is Not a Phase →08
Trust should not be assumed.
It should be justified.
When AI assumes responsibility or influences decisions, trust should be more than a feeling. Let us make visible what that trust is based on.
Discuss with scip →