Development record · Version 0.1

Open Issues Register

The unresolved questions that remain visible during public development, organised by subject and validation priority.

Version: 0.1
Date: 16 August 2026
Purpose: make unresolved normative, methodological, legal and validation questions explicit rather than convert them silently into requirements

Priority means: High — material to interpretation or safe operational use and an early v1.0 input; Medium — important for consistency or wider adoption; Low — useful refinement or future consideration.

1. Scope, qualification and engagement terms#

Issue ID Subject Why it remains open Impact on v0.1 Evidence or validation needed Priority Possible next step
OI-01 Engagement-level conformity wording Self-declaration can be mistaken for certification, assurance or owner approval; no conformity-assessment project was undertaken. REQ-01 uses a narrow claim and forbids partial/provider-level claims. User comprehension and legal review of specimen claim language across jurisdictions. High Test alternative wordings with boards, practitioners and assurance professionals before v1.0. Any third-party scheme requires a separate authorised project.
OI-02 Minimum pre-acceptance information Enough information is needed to qualify, but a fixed list may exclude urgent or confidential cases. REQ-02 states factors, not a documentary threshold. Pilots comparing accepted, declined and reclassified engagements. Medium Develop a non-normative qualification evidence checklist and validate it.
OI-03 Emergency or compressed-timing engagements Urgency can be real, but it can also be used to bypass evidence and challenge. Urgency cannot remove essential conditions; it may narrow the conclusion. Case studies in crisis, transaction and regulatory settings. High Pilot explicit “time-limited basis” conclusions and test user understanding.
OI-04 Meaningful remaining decision latitude A nominal future gate may conceal practical or path-dependent commitment. REQ-02 requires meaningful ability to act, but no metric exists. Comparative cases on contractual, financial, political and operational lock-in. High Define observable indicators of alter, condition, stage, defer and reject capacity.
OI-05 Intended-user participation or consent Intended users may not commission the work or agree the question; mandatory consent may be impossible. REQ-03 requires identification, not consent. Governance/legal review of engagement roles and reliance expectations. Low Test whether acknowledgement, rather than consent, improves use.
OI-06 Privileged, restricted or inaccessible information Access may be legally restricted while the information is material; privilege should not be waived by the Standard. Limitations must affect the basis/conclusion; no access power is created. Jurisdictional legal review and privilege-sensitive case design. High Develop legally reviewed patterns for reliance, clean teams, summaries and inability to conclude.
OI-07 Threshold for material scope change Too-low threshold creates administration; too-high permits scope laundering. REQ-03 requires material changes to be requalified. Practice data on question evolution and conclusion impact. Medium Record and classify scope changes during pilots; derive examples.
OI-08 Mandate for internal practitioners Charters, reporting lines and management influence vary substantially. v0.1 uses a functional mandate test. Comparative validation with internal audit, risk and strategy functions. Medium Pilot internal-function cases and publish neutral examples.
OI-09 Distribution, legal reliance and third-party rights Evidential use, permitted distribution and legal duty are jurisdiction/contract specific. The Standard creates no distribution or legal reliance rights. External legal research in priority jurisdictions. High Obtain jurisdiction-specific counsel before operational report templates are used.

2. Competence and independence#

Issue ID Subject Why it remains open Impact on v0.1 Evidence or validation needed Priority Possible next step
OI-10 Sector-specific competence indicators A generic standard cannot specify competence for every technical, legal, financial or policy question. REQ-04 uses an engagement-specific collective test. Sector panels and analysis of failure cases. Medium Develop, but do not yet normatively adopt, sector competence profiles.
OI-11 Competence of the output authoriser A multidisciplinary conclusion may exceed any one person's technical depth; ceremonial senior sign-off is a risk. REQ-13 requires sufficient understanding, not mastery of every specialty. Observation of authorisation reviews and expert interviews; inter-rater testing. High Define evidence that an authoriser understood dispositive specialist judgments.
OI-12 Prior authorship and cooling-off Prior work creates self-review/advocacy risk, but a universal period is unsupported and expertise can be scarce. REQ-05 applies a threat test; no automatic period. Comparative professional rules and engagement pilots. High Test a matrix of authorship materiality, elapsed time, personnel separation and independent rework.
OI-13 Implementation-service interests Expected downstream revenue can impair independence without a signed contingency. REQ-05 captures the threat; REC-01 recommends separation. Firm-structure, remuneration and behavioural evidence across providers. High Validate when another provider is required rather than safeguards within one network.
OI-14 Organisational independence of internal functions External status is not sufficient and internal status is not disqualifying; objective thresholds are immature. Functional independence test only. Case comparisons by reporting line, prior role, budget and release authority. High Develop observable internal-independence indicators without copying internal-audit charters wholesale.
OI-15 Reasonable and informed observer test Professional codes use similar tests, but legal interpretation and cultural expectations vary. v0.1 uses reasonable appearance without making a legal test. Comparative ethics/legal analysis and user research. Low Add interpretive examples after validation.

3. Evidence, reliance and uncertainty#

Issue ID Subject Why it remains open Impact on v0.1 Evidence or validation needed Priority Possible next step
OI-16 Operational threshold for an adequate basis No universal non-assurance evidence threshold exists for heterogeneous decision questions. REQ-06 ties adequacy to material aspects and conclusion strength; judgment remains. Independent re-review of completed cases; inter-rater agreement; false-positive/false-negative analysis. High Make this the first practice-validation workstream.
OI-17 Confidential or non-shareable evidence Reviewers and intended users may be unable to inspect material support. REQ-06/15 require traceability subject to law; limitation can constrain conclusion. Secure-review models and legal analysis. High Test sealed evidence summaries, independent verification statements and access-tier designs.
OI-18 Source quality in novel markets and technologies Relevant evidence may be sparse, vendor-controlled or rapidly obsolete. REQ-06 requires explicit provenance/uncertainty rather than rejecting novelty. Sector case studies and later outcome-independent re-review. Medium Create examples of bounded conclusions under sparse evidence.
OI-19 Depth of third-party provider objectivity assessment ISA/IESBA models concern defined expert roles; client reports and market data vary. REL-01 says “to the extent relevant”. Reliance mapping across legal, commercial, technical and data sources. Medium Pilot a material-reliance classification and required response levels.
OI-20 Proprietary expert or model opacity Intellectual property can prevent method inspection, but dispositive opaque work may not support a defensible conclusion. REQ-07/14 require evaluability or limited reliance. Controlled tests of black-box work, independent benchmarks and error cases. High Define minimum evaluability outcomes rather than disclosure of proprietary detail.
OI-21 Use of another assurance report An assurance report may support some claims, but subject, criteria, users, date and assurance level may not fit the decision use. REL-01 requires fitness assessment and preserves the original report boundary. Assurance-professional guidance and worked cases. Medium Produce non-normative examples of appropriate and inappropriate use.
OI-22 Deep or irreducible uncertainty Probabilities or stable scenarios may be unavailable; ordinary sensitivity analysis can mislead. REQ-08 permits qualitative treatment and requires bounded conclusions. Decision-science and sector expertise on robust/staged decisions. High Pilot conclusion wording for ambiguity and deep uncertainty.
OI-23 When sensitivity analysis is necessary Universal modelling is disproportionate; omitting it for a key assumption can be indefensible. REQ-08 uses a necessary-and-proportionate test. Retrospective review of which assumptions drove conclusion reversals. Medium Derive trigger examples by decision type.
OI-24 Sector-specific quantification expectations Cost, safety, valuation and policy decisions have different valid methods and professional rules. v0.1 has no universal quantification requirement. Sector standards mapping and expert panels. Medium Consider future sector guidance only after the core Standard is validated.

4. Challenge and dissent#

Issue ID Subject Why it remains open Impact on v0.1 Evidence or validation needed Priority Possible next step
OI-25 Stopping rule and minimum evidence for genuine challenge Unbounded search is wasteful; token challenge is cosmetic. No validated universal metric exists. REQ-09 uses materiality, plausibility and proportionality. Blind review of engagement files; comparison with independent challenger judgments. High Make this the second practice-validation workstream.
OI-26 Access to unshared or suppressed information Group evidence shows a failure risk, but practitioners cannot know every undisclosed fact. Mandate and challenge require reasonable access/search, not completeness. Interviews, whistleblowing/escalation design and case studies. Medium Test confidential issue-solicitation routes in higher-risk cases.
OI-27 Gaming-resistant challenge observability Counts of questions, workshops or alternatives are easily gamed. v0.1 assesses material reasoning/disposition. Adversarial scoring trials on real and synthetic case files. High Develop qualitative review anchors and test inter-rater reliability.
OI-28 Dissent disclosure versus release blocking Audit requires resolution in defined contexts; transparent unresolved dissent may be more honest here. Threshold is unvalidated. REQ-10 requires disclosure if user understanding is affected, otherwise withholding/withdrawal where disclosure is impossible. Practitioner/user experiments with dissent scenarios and report variants. High Make this the third practice-validation workstream.
OI-29 Privileged, confidential or personal content in dissent The substance may be material but disclosure can be unlawful or harmful. REQ-10 permits in-substance communication and otherwise withholding. Legal review and ethics guidance. High Develop anonymised/substance-only examples and escalation rules.
OI-30 Protection and independence of dissent escalation A formal channel does not prevent retaliation, self-censorship or removal from the team. v0.1 requires a means and non-suppression but no employment-governance system. Organisational research, ethics hotlines and professional quality models. High Test independent consultation/escalation options; keep broader whistleblowing governance out of scope.
OI-31 Solo-practitioner treatment No internal dissent exists, and a separate reviewer may be unavailable. Self-challenge can be insufficient. CHAL-01 applies; REC-02 may justify external review. Solo-practice pilots and external consultation models. Medium Define heightened triggers for external consultation without making it universal.

5. Conclusions, reporting and authority#

Issue ID Subject Why it remains open Impact on v0.1 Evidence or validation needed Priority Possible next step
OI-32 Standard conclusion taxonomy Common labels aid comparability but risk assurance confusion and oversimplification. v0.1 permits flexible adverse, supportive, conditional and no-conclusion forms. User comprehension and cross-sector report review. Medium Test whether a small descriptive taxonomy improves interpretation without becoming a rating.
OI-33 Prominence in short, slide-based or digital reports Presence of a limitation does not make it visible; format effects vary. REQ-11/12 require proportionate prominence and whole-report clarity. Board/executive comprehension studies using report variants. High Test placement, hierarchy and summary/full-report consistency.
OI-34 Updating after report date New information can make a point-in-time conclusion obsolete; an open-ended duty is impractical. Guidance states no implied updating duty. Legal/professional review and staged-decision practice. Medium Develop contractual update triggers and report-date warnings outside the core requirement.
OI-35 Recommendations versus conclusions A second opinion may need practical conditions, but detailed solution design can become consultancy. Category boundary permits decision implications, not management ownership. Review of combined advisory/second-opinion engagements. Medium Develop examples separating evaluative conditions from implementation advice.
OI-36 Naming the output authoriser Personal naming increases accountability; role-only identification may be necessary for security or internal practice. REQ-12 permits identity or accountable role. User trust/comprehension and legal/employment review. Low Test both forms; retain traceable internal identity regardless of publication.
OI-37 Output authorisation in internal hierarchies Senior approvers may be distant; engagement leads may lack release authority. REQ-13 is functional and permits role variation. Internal-function pilots and governance mapping. Medium Define examples of acceptable authorisation chains without prescribed titles.

6. Technology and documentation#

Issue ID Subject Why it remains open Impact on v0.1 Evidence or validation needed Priority Possible next step
OI-38 Report disclosure threshold for tool use Listing ordinary tools is noise; undisclosed material AI/model use can mislead. REQ-14 requires internal traceability and disclosure of material limitations, not every tool name. User studies and model-risk cases. Medium Test a material-contribution and material-limitation disclosure decision tree.
OI-39 Evidence of meaningful human review A signature and logs do not prove understanding or override capability. REQ-13/14 specify substantive review but no metric. Observed review sessions, error-seeding tests and reviewer interviews. High Validate through seeded-error pilots in material automated analyses.
OI-40 Changing AI law and standards EU AI Act phases, NIST revision and professional standards will evolve. v0.1 stays technology-neutral and defers to applicable law. Scheduled primary-source monitoring. Medium Review the Source Register at least annually and before v1.0.
OI-41 Minimum retention period Legal, professional, contractual and risk needs differ. REQ-15 requires proportionate retention under applicable obligations but no period. Multi-jurisdiction legal/professional review. Medium Develop jurisdiction/sector guidance rather than a universal core period.
OI-42 Access to documentation for independent review Traceability is difficult to test if records are private; access can conflict with confidentiality. No third-party inspection right is created. Governance, privilege and independent-review pilots. Medium Test controlled reviewer access and redacted evidence maps.
OI-43 Documentation confidentiality and privilege Detailed challenge records can expose sensitive strategy, personal data or legal advice. Applicable obligations prevail; insufficient lawful traceability may limit conformity. Legal/security/privacy review. High Create data-classification and privilege-handling guidance outside the core Standard.
OI-44 Dynamic or continuously updated reports Versioning, authorisation and point-in-time boundaries are unclear in live systems. v0.1 requires a controlled authorised version. Digital-report pilots and records-management analysis. Low Treat each authorised state as a dated release until validated otherwise.

7. Proportionality, withdrawal and future development#

Issue ID Subject Why it remains open Impact on v0.1 Evidence or validation needed Priority Possible next step
OI-45 Calibration of proportionality factors Factors are defensible but their relative weight and effect on work are unvalidated. P-02 lists factors and a non-negotiable floor. Cross-case effort/outcome-independent quality analysis. Medium Collect planning rationales and independent file-review results in pilots.
OI-46 Time-constrained decision treatment Time can be a legitimate boundary or a sponsor-created excuse; there is no safe universal minimum. REQ-02, 06, 11 require a basis and honest disposition despite urgency. Crisis/transaction case pilots and user tests of limited conclusions. High Combine with OI-03 in early validation.
OI-47 Sector annexes or higher floors Sector variation may justify additional rules, but premature annexes can fragment the category. v0.1 remains cross-sector. Sector-source mapping, pilots and stakeholder review. Low Consider only after core inter-rater and practice validation.
OI-48 Withdrawal rights and communication duties Contract, employment, regulation and public-interest duties differ. REQ-11 requires an appropriate disposition subject to law. Jurisdictional legal and professional review. High Obtain legal input before engagement terms and operational procedures are issued.
OI-49 Client use of drafts or withdrawn reports A sponsor can circulate an unauthorised favourable draft despite withholding. REQ-12/13/15 require controlled status and release, but enforcement is external. Contractual controls, records practice and incident cases. High Develop clear watermark/status and contractual-use provisions as implementation guidance.
OI-50 Terminology for limited conclusions “Qualified” and “modified” imply audit; “conditional” can be mistaken for approval subject to conditions. v0.1 uses explicit plain-language boundaries. User comprehension across professions. Medium Test report examples and refine definitions for v1.0.
OI-51 Threshold for separate pre-release review Audit models are mature but category-wide benefit/cost is unknown. REC-02 remains non-mandatory. Comparative file reviews with/without separate review. Medium Pilot high-risk triggers and measure material corrections, not reviewer completion.
OI-52 Validity of numerical confidence communication No general calibration exists for heterogeneous one-off decisions. Numerical score is not required and cannot substitute for boundaries. Decision-type-specific calibration and user-interpretation studies. Low Do not operationalise in v0.1; revisit only with independently validated methodology.
OI-53 Ordinary-language assurance perception Even with disclaimers, “independent opinion” and “Standard” can create assurance-like expectations. v0.1 includes repeated non-assurance boundaries. User testing, assurance-professional review and legal analysis of report presentation. High Test full specimen reports, not disclaimers in isolation, before public operational claims.

8. Priority validation sequence#

The first three practice-validation programmes should be:

  1. Adequate basis and conclusion calibration — OI-16, linked to OI-17–24 and REQ-06–08.
  2. Genuine challenge observability and stopping — OI-25 and OI-27, linked to REQ-09.
  3. Material dissent disposition — OI-28–30, linked to REQ-10 and REQ-13.

Meaningful human review (OI-39) is the next priority where material automated analysis is actually used; it is not made the first universal validation stream because REQ-14 is conditional.