HIA exists for the space that governance design cannot reach. Most AI governance programmes produce frameworks that define what should happen when a system behaves unexpectedly. They rarely determine whether the people responsible for acting on those frameworks can do so, under time pressure, with incomplete information, and against authority structures that were not designed with this moment in mind.
Our work is practitioner-led. Every assessment is conducted by advisors with direct operational experience in AI systems environments, not analysts evaluating your governance against a compliance checklist. We evaluate against operational reality.
- Every engagement begins with a focused, confidential conversation to establish context and determine whether HIA is a fit for the situation.
- A written scope follows, defining what will be assessed, how, and what the engagement will produce. No generic frameworks, no standardised outputs.
- Assessments are conducted by senior advisors with direct operational experience. Findings are accurate, not reassuring.
- Findings are presented directly to the board or executive team: clear, prioritised, and specific to the conditions of your organisation.
Three primary service areas. Each is contained, with clear scope and defined outcomes.
- Assessment of decision quality and authority clarity under pressure.
- Review of information adequacy: what the board actually receives versus what it needs to act.
- Independent view not shaped by those who designed the governance framework being assessed.
- Written capability assessment report
- Decision authority gap map
- Prioritised remediation recommendations
- Board-ready executive summary
- Project-based or ongoing retainer
- Direct board and executive access required
- 6 to 12 weeks typical project engagement
- Remote or in-person
Focus: decision clarity. A board that cannot make effective decisions when governance is tested is not a safeguard.
Request outline- Mapping of formal and informal escalation paths from system event to human response.
- Assessment of override authority: who can act, under what conditions, with what information.
- Identification of structural gaps and decision latency that would prevent timely intervention.
- Escalation architecture document
- Decision rights map
- Structural gap report
- Prioritised remediation plan
- 4 to 8 weeks
- New and existing AI deployments
- Tailored to your operational context
- Direct presentation to accountable executive
Focus: the mechanics of real-time decision-making when a system produces an output that no one planned for.
Request outline- Realistic AI system scenarios under time pressure, with incomplete information and real authority constraints.
- Diagnostic of where decision-making breaks down in practice, not in theory.
- Honest account of what would need to change for intervention capability to be reliable.
- Simulation debrief report
- Intervention capability assessment
- Prioritised action list
- Board-level summary
- 2 to 3 weeks scenario design, then a half to full-day session
- Up to 12 executive participants
- Remote or in-person
- Can stand alone or follow diagnostic
Focus: surface the discovery before the event, not after consequences have compounded.
Request outlineThe right service depends on the specific question you are holding. If it is unclear, the first step is a confidential conversation.
If you recognise your situation in any of the service areas above, the first step is a short, confidential conversation to clarify what is actually at stake and which service would address it. Please include your role, the situation prompting the enquiry, and any relevant timelines.