Trustworthiness Optimisation Process
TOP
AI is transforming high-impact sectors, from healthcare to maritime operations. But while principles like fairness, transparency, and accountability are widely recognised, putting them into practice remains a challenge.
The THEMIS Trustworthiness Optimisation Process (TOP) bridges this gap. It is a practical, risk-based methodology for assessing and improving the trustworthiness of AI systems across their lifecycle.
Turning AI Principles into Real-World Trust
What is TOP?
TOP is a four-stage process that helps organisations move from abstract AI ethics to concrete, auditable action. It combines:
​
-
AI lifecycle thinking
-
Risk management
-
Human oversight
-
Trustworthiness metrics
-
Decision-making tools
Unlike single-focus approaches, TOP enables organisations to identify risks, apply solutions, and balance trade-offs across multiple trust dimensions
Why it Matters
Trustworthy AI doesn’t happen through compliance alone. It requires clear methods, structured decisions, and continuous evaluation.
TOP helps organisations to:
​
-
Identify and prioritise AI risks
-
Apply targeted mitigation strategies
-
Document decisions transparently
-
Involve stakeholders
-
Balance trade-offs (e.g. fairness vs accuracy)
​
​
How it Works
TOP follows a simple but powerful structure:
-
Understand the AI system and its context
-
Assess trustworthiness risks
-
Improve using targeted methods
-
Optimise by balancing trade-offs
​
Within THEMIS, TOP is applied in real-world settings across healthcare, media, and maritime sectors, where trust is critical. It enables organisations to move from AI adoption to trustworthy, risk-aware deployment.