Get the Standard
The TRUST-AI Open™ standard is the Responsible AI Standard that assesses both organisational behaviours and system design. It’s open, actionable, and ready to implement.
Designed for real-world AI teams, TRUST-AI Open™ includes five core domains: Alignment, Data, Implementation, Impact, and Adaptation. It bridges the gap between ethical intent and operational practice, helping teams move from principles to measurable progress.
The TRUST-AI Open™ Standard will be released publicly in July 2025. Sign up to be notified when it becomes available for purchase.
What makes TRUST-AI Open™ different?
Dual-layered: Addresses both organisational culture and AI system design
Behavioural focus: Goes beyond compliance to assess ethical maturity
Built for teams: Includes readiness prompts, narrative scoring and facilitation cues
Verification-ready: Forms the foundation for our tiered certification model
The Five Key Elements of TRUST-AI Open™
1. Purpose Alignment
Why are we building this—and for whom?
Clear articulation of intent and value
Alignment with human needs, rights, and context
Avoidance of unintended or unethical outcomes
“Without purpose, AI becomes an accelerant of whatever system it's in, good or bad.”
3. Transparency & Accountability
Can we see it, question it, and improve it?
Explainability of models and decisions
Traceability of data and system logic
Human accountability for automated outcomes
“Black box = red flag. If no one is answerable, no one is responsible.”
2. Fairness & Inclusion
Who benefits—and who bears the cost?
Bias mitigation in data, design, and deployment
Inclusive access across groups and geographies
Avoidance of harm to vulnerable populations
“If it only works for some, it doesn’t work responsibly”
4. Safety & Security
Does it protect people from harm?
Robustness to misuse, manipulation, or failure
Data integrity and cyber resilience
Protocols for escalation, override, and redress
“Trust dies the moment something unsafe slips through because the model said so.”
5. Human Agency & Oversight
Do humans stay in the loop—and in charge?
Oversight that empowers people, not bypasses them
Feedback loops, contestability, and continuous improvement
Cultural readiness and capability to work with AI, not under it
“Responsible AI isn’t about rules—it’s about relationships.”
TRUST-AI Open™ Comparison
While frameworks like NIST, OECD, and UNESCO define essential AI principles, the TRUST-AI Open™ Standard goes further—operationalising them with practical tools, behavioural scoring, and tiered certification built for real teams.
| Pillar / Focus Area | NIST AI RMF | OECD Principles | UNESCO Recommendation | TRUST-AI Open™ |
|---|---|---|---|---|
| Implementation Readiness | Abstract controls and mapping tools | Policy-focused principles | Governmental guidance and values | Includes elements from all other standards, plus ready-to-use tools, prompts, and certification pathways |
| Purpose Alignment | Context mapping, governance roles | Well-being, inclusive growth | Accountability, awareness | Includes elements from all other standards, plus operationalises purpose in daily work; prevents drift |
| Fairness & Inclusion | Bias checks, fairness indicators | Non-discrimination, inclusivity | Collaboration, equity | Includes elements from all other standards, plus participatory design and population-specific bias auditing |
| Transparency & Accountability | Explainability, traceability | Transparency, accountability | Public understanding, redress | Includes elements from all other standards, plus explainability for users, logging, safe internal challenge |
| Safety & Redress | Risk scanning, recovery planning | Security, override options | Oversight, resilience | Includes elements from all other standards, plus redress tools, human override, lifecycle design |
| Human Agency & Oversight | Governance maturity models | Human empowerment, cooperation | Participation, governance feedback | Includes elements from all other standards, plus behavioural scoring and evolving governance built in |