Data quality
Is the data accurate, complete, consistent and timely?
HUMAN-CENTRIC INFORMATION INTEGRITY FRAMEWORK
An Enterprise Information Integrity Management System for establishing trust and confidence in the information organisations rely upon.
Explore HCIIFHaving information does not automatically mean it is appropriate to act on it.
Quality matters. So do provenance, context, sensitivity, accountability and human judgement. HCIIF brings these considerations into a coherent enterprise approach to information integrity.
It asks a practical question: is this information sufficiently trusted for its intended purpose, risk and consequence?
Data is the starting point. Trusted information is the outcome.
Organisations invest in data quality, data governance and AI governance. These disciplines are essential—but data is interpreted, combined, transformed, shared and relied upon in a wider organisational context.
HCIIF builds on those foundations and extends them into Enterprise Information Integrity: establishing whether information is sufficiently trusted for its intended purpose, risk and consequence.
Is the data accurate, complete, consistent and timely?
Is the data appropriately owned, controlled and managed?
What confidence is justified when data informs people, processes and AI?
Is the resulting information sufficient for the intended decision or action?
A human-centric, risk-based approach that connects evidence, assurance and accountable use across the information lifecycle.
Establish what information is being relied upon, where it comes from, its context and the consequences of using it.
Assess the evidence, limitations and conditions that justify confidence. Communicate that basis through the Information Integrity Passport.
Keep confidence under review as information, circumstances and intended uses change, with clear accountability and proportionate assurance.
“The objective is not perfect information.
It is sufficiently trusted information.”
Make the evidence and limitations behind information-led judgements visible.
Connect information risk, ownership and assurance to the purpose of use.
Strengthen the information foundations on which AI-enabled processes and outcomes depend.
Deepak Sadasivan
Currently undergoing peer review
The architecture, principles and operating model behind HCIIF.
Across 20 chapters, the Enterprise Edition explores how organisations can understand, assess, communicate and sustain confidence in enterprise information.
The publication edition will be made available here after peer review and final revisions.
Enquire about the frameworkIn development
A practical companion for organisations applying the Enterprise Edition.
The Practitioner Guide will translate the framework’s architecture and principles into implementation guidance for practitioners. It is intended to support proportionate application across different organisational contexts—not prescribe a single operating model.
Scoping, information-asset identification, confidence assessment and Information Integrity Passports.
Governance roles, lifecycle controls, assurance and integration with existing enterprise capabilities.
Adoption, maturity, continual improvement and evidence-led measurement.
The guide is being developed alongside peer review and practical validation. Its scope and publication timing will be confirmed when that work is sufficiently mature.
From the information trust challenge to a sustained enterprise capability. Browse the chapter summaries to see how the framework fits together.
Explores why access to information is not enough for confident action. Establishes the need to understand its origins, evidence, limitations and suitability for the decision at hand.
Defines information integrity in its enterprise context, bringing together meaning, provenance, quality, authority and purpose. Explains why technically accurate information can still be unsuitable for a particular use.
Introduces HCIIF as a human-centric management system for establishing and sustaining justified confidence. Connects information integrity to organisational purpose, professional judgement, accountability and consequence.
Sets out how leadership, governance, planning, operation, assurance and improvement work together. Explains how to embed HCIIF proportionately within existing organisational structures and capabilities.
Explains the evidence-led methodology for assessing confidence for a defined purpose. Makes strengths, limitations, contradictions and professional judgement visible through an explainable Confidence Basis.
Introduces the governed, traceable representation of an information asset’s integrity context. Shows how purpose, provenance, confidence, rights, caveats and history remain available to authorised users.
Explains how confidence is monitored, challenged and maintained as information and circumstances change. Connects reassessment, emerging gaps, drift, intervention and escalation to proportionate organisational action.
Identifies the information that warrants deliberate integrity management because of its value, sensitivity, dependencies or consequences. Helps organisations focus assessment and assurance where reliance matters most.
Follows confidence through information creation, use, transformation, transfer and retirement. Explains when scheduled reviews, events and exceptions should trigger reassessment or action.
Distinguishes justified confidence from permission to access, use or share information. Examines how rights, restrictions and context accompany information, while recipients remain accountable for their own reliance.
Explains how ongoing assurance evidence becomes insight into patterns, trends and emerging weaknesses. Connects that intelligence to proportionate governance responses and earlier attention to material risks.
Describes the technology-neutral architecture that brings confidence context into existing systems, workflows and AI capabilities. Makes governance information available where people and technologies rely upon information.
Sets out responsibilities for ownership, assessment, decisions, acceptance of uncertainty and escalation. Separates evidence-led confidence judgements from the authority to act, within a federated operating model.
Shows how to begin with a defined organisational need, assess existing capability and prioritise improvements. Treats maturity as demonstrated effectiveness and sustainable practice, supported by learning and proportionate adoption.
Explains how to measure changes in capability, reliance and organisational outcomes. Connects those changes to benefits and enterprise value while distinguishing HCIIF’s contribution from unsupported claims of causation.
Explores how leaders use integrity evidence to govern consequential decisions, investment and organisational direction. Makes rights, caveats, confidence, consequences and decision authority explicit.
Examines how to preserve integrity context through migrations, organisational change and disruption. Explains how to operate under degraded conditions and re-establish justified confidence during recovery.
Applies HCIIF before, through and after AI reliance, including source information and derived outputs. Explains why confidence in inputs does not automatically establish confidence in AI-generated information.
Brings the capabilities together as a sustained enterprise management system. Distinguishes implementation, effectiveness and maturity, and examines what is needed to keep HCIIF operating within an agreed scope.
Draws together the framework’s implications for information moving across organisations, suppliers, sectors and national boundaries. Returns to the central question: what confidence is justified, under what conditions, and for which intended use?
Summaries reflect the Enterprise Edition currently undergoing peer review.
FROM PRINCIPLES TO RELIANCE
Three illustrative scenarios show how information integrity affects decisions across commercial, international and healthcare settings.
Illustrative scenario
A service launch. Three critical suppliers. One executive briefing that says everyone is ready.
An organisation is preparing to launch a new service. A consolidated report marks its critical suppliers as ready. An AI assistant summarises that report for the launch meeting: “All critical suppliers are ready for launch.”
But the consolidation has lost important qualifications. One supplier’s evidence relates to an earlier configuration. Another supplier’s readiness is conditional on an unresolved dependency. The executive summary presents a stronger conclusion than the evidence supports.
Recent acceptance evidence covers the intended launch scope. Its source and accountable owner are identifiable.
The evidence predates a material configuration change. Applicability to the planned launch needs reassessment.
The supplier’s report depends on an outstanding integration check. That condition is missing from the consolidated summary.
The consolidated readiness report is identified as an Enterprise Information Asset, linked to its supporting supplier evidence. The intended use is the launch decision for a defined service scope and configuration. Required Confidence is established in that context, with attention to the consequences of disruption.
ECAM examines provenance, currency, scope, contradictions and limitations. The available evidence does not justify the blanket claim that all three suppliers are ready. This is a gap in the basis for reliance; it does not, by itself, prove that a supplier is unready.
The leader considers whether to delay, narrow the launch or proceed with defined conditions. That judgement includes wider operational and risk evidence. Any permitted acceptance of reliance below Required Confidence is recorded within the relevant authority; it does not remove the Confidence Gap or override rights and restrictions.
The briefing is corrected to retain the qualifications. Existing recipients are notified where their reliance may be affected. AI-generated summaries are assessed for their intended use; confidence in source material does not automatically transfer to a derived conclusion. New supplier evidence triggers review of the report, Passport and affected briefings.
The launch discussion moves from an unexplained “ready” label to a visible basis for judgement: what is supported, what remains uncertain, what conditions apply and who can decide.
The intended benefit is earlier, better-informed action on uncertainty. No pilot results, measured savings or validated outcomes are claimed.
Illustrative scenario
The same assessment crosses a border. Its intended use changes. Does the original basis for reliance still apply?
An organisation shares a regional service-disruption assessment with an overseas partner to support contingency planning. It describes likely disruption over the coming month, with qualifications about incomplete source coverage and the period assessed.
The recipient later proposes using it to reroute a live service that day. A shortened internal briefing omits the source limitations and onward-sharing conditions. The original assessment remains unchanged, but the new decision needs more current and specific evidence.
A regional assessment supports planning for possible disruption over a defined period. Its Confidence Basis is tied to that purpose and scope.
The recipient needs evidence about a specific service, location and time. That need is not automatically met by the broader assessment.
Incomplete source coverage, the assessment period and onward-sharing restrictions are absent from the briefing reaching the decision-maker.
The recipient traces the briefing back to the original assessment and its supporting context. The information owner clarifies the purpose, assessment period, limitations and conditions of the original sharing. The shortened briefing is treated as a derived information product whose suitability also requires attention.
Authorised roles check whether the proposed use and any onward disclosure are permitted under the applicable agreements and requirements. Separately, the assessor considers whether the existing Confidence Basis applies to the operational decision and meets Required Confidence. Reliable information does not create permission, and permission does not establish sufficient confidence.
The recipient can seek current evidence, clarify sharing permissions, retain the assessment for its original planning purpose or escalate the operational decision. If action under uncertainty is permitted, the relevant authority records its rationale and conditions. Accepting uncertainty does not override a restriction on use or disclosure.
The missing qualifications are restored to the internal briefing. If later evidence changes the assessment, material updates or recall notices reach authorised recipients and affected downstream users. Recipients review their reliance and derived products rather than assuming the previous position remains valid.
The recipient can distinguish what the sender’s assessment supports, what the new decision requires and what use is permitted. Context becomes part of the exchange, not something left behind when the information moves.
This illustrative scenario demonstrates the intended use of HCIIF. It does not describe a real operation or claim validated outcomes.
Illustrative scenario
Hospital, primary care and community services must rely on the same discharge information—but a late medication change has not reached every record.
A patient is being discharged from hospital with follow-up care provided by a GP and a community team. The electronic discharge summary contains the expected medication list, care instructions and follow-up arrangements.
Shortly before discharge, one medication is changed following a clinical review. The change appears in the prescribing system but not in the version of the discharge summary prepared for onward transfer. An AI-enabled summarisation tool reproduces the earlier list without identifying the difference between the source records.
The document contains the expected fields and an identifiable author, but its medication content predates the final review.
A more recent entry records an amended medication decision, creating a material contradiction requiring resolution.
The generated summary repeats the earlier list without exposing its source version, timing or the conflicting record.
The discharge information is identified as an Enterprise Information Asset supporting medication reconciliation, follow-up care and communication across organisational boundaries. Required Confidence reflects the potential consequence of relying on an incorrect or outdated medication position.
ECAM examines provenance, timing, authority and consistency across the relevant records. The conflicting versions mean that the current discharge summary does not yet provide a sufficient basis for reliance on the medication list. This identifies a Confidence Gap for resolution by an appropriately authorised healthcare professional.
An authorised professional reconciles the records and confirms the current medication position. The discharge summary and its integrity context are updated, affected recipients are notified, and the correction is traceable to the source decision rather than silently overwriting the earlier version.
The AI-generated summary is treated as a separate information product. Its source, currency and intended use are assessed, and it is regenerated or corrected after reconciliation. Confidence in a source record is not assumed to transfer automatically to a derived output.
Teams can distinguish a document that looks complete from information that is sufficiently current, consistent and authoritative for the intended care decision. The contradiction becomes visible early enough for accountable human resolution.
This illustrative scenario demonstrates the intended use of HCIIF. It does not provide clinical advice, describe a real patient or claim validated outcomes.
START A CONVERSATION
For enquiries about HCIIF, peer review, collaboration or potential pilots, please get in touch.
Contact contact@hciif.com