Independent perspective
Independent review that asks difficult questions.
Independent AI governance challenge
AI governance should survive contact with real work.
Providing independent challenge on AI governance, accountability, decision ownership and oversight.
Stronger accountability
Clearer decision ownership and oversight in practice.
Evidence that stands up
Governance and evidence that can be explained, defended and reconstructed.
Insights
Where governance meets real-world accountability.
2 June 2026
Boeing 737 MAX and the Limits of “The Human Can Intervene”
"The human can intervene" is one of the most reassuring things an organisation can say about a safety-critical system.
30 May 2026
The Post Office Had Humans in the Loop Too
"Human-in-the-loop" is one of the most reassuring phrases in AI governance. It suggests there is still a person in the chain.
Additional observations and commentary are published on LinkedIn.
View LinkedIn →What independent review can surface.
Governance written from the inside often reflects what an organisation intends to happen. Independent review tests whether accountability, oversight and evidence remain clear when AI use becomes part of real work.
Hidden accountability gaps
Oversight that is too vague to test
Evidence that may not exist when needed
Governance that is documented but difficult to evidence
Independent Review
Findings are shaped by the context, documentation, workflows and decision processes available for review. The scope depends on the question being examined and the information available.
Specific questions
A review can examine a specific decision, workflow, governance statement or area of concern.
Broader arrangements
A broader review can examine how governance, oversight and accountability operate across a defined area of activity.
Evidence focus
The focus is evidence: what exists, what is missing, and what can be reasonably supported.
The objective is to understand how governance operates in practice.
About Kim Stevens
Kim Stevens is an independent advisor focused on accountability, decision ownership and oversight in AI-enabled work.
Her background includes clinical research, project management and AI governance. Current work focuses on accountability in operational settings, governance language, evidential reconstruction and Behavioural Accountability & Traceability (BAT).
She writes and researches questions relating to how governance operates in practice, particularly where policy language, human oversight and accountability become difficult to evidence.
Contact
For research, independent review or a specific accountability question, email Kim Stevens directly or leave your details here.
Email Kim Stevens