Working Draft — content version 1.4.0 · review package 1.2 · not approved content
SRCF / framework / domain I / I.4
Domain I · Area I-CA2 · Responsible AI use, grounding and model-risk control

I.4Use AI for research, retrieval and source discovery

Working Draft version 1.1

Purpose and scope

This unit covers the capability to use AI-assisted search, retrieval, summarisation and question-answering tools to support sustainability reporting research. It includes defining the research question, source hierarchy, retrieval scope, dates and versions; distinguishing discovered, retrieved and authoritative sources; verifying citations and quotations; maintaining a research log; and protecting against fabricated, outdated, incomplete or context-lost information.

Applied competency statement
Can use AI to support research and source discovery while independently verifying source authority, currency, version, scope, quotation, citation and relevance before using information in reporting work.
Boundary and escalation
This unit owns AI-specific source discovery, retrieval and research documentation. A.2 remains the canonical owner of source authority, hierarchy, version, provenance and citation and I.6 owns verification of AI-assisted outputs. AI retrieval does not make a source authoritative or replace legal, technical, standards or specialist research.
Key quality risks
AI response treated as a source; fabricated citations or inaccessible links; outdated versions used; unofficial commentary substituted for authoritative requirements; quoted language not present in the source; search result snippets used without opening the document; important exceptions or context omitted by summarisation; different jurisdictions or dates merged; confidential questions entered into unauthorised tools; and research output reused without recording the model, date or verification.
Required knowledge · 6
I.4-K01
Understands search, retrieval, semantic search, summarisation, extraction, question answering, retrieval-augmented generation and the difference between model knowledge and retrieved content.
Knowledge Type: AI-assisted research modes · Normative Weight: Core
I.4-K02
Understands primary and secondary sources, official and unofficial publications, effective and superseded versions, jurisdiction, date, issuer and source purpose.
Knowledge Type: Source authority and hierarchy · Normative Weight: Core
I.4-K03
Understands incomplete retrieval, chunking, context-window limits, ranking, inaccessible attachments, image or table loss, translation and the risk of merging different sources or periods.
Knowledge Type: Retrieval and context risk · Normative Weight: Core
I.4-K04
Understands source opening, exact-location checking, quotation limits, citation accuracy, document identity, URL or identifier stability and version retention.
Knowledge Type: Citation and quotation verification · Normative Weight: Core
I.4-K05
Understands research questions, search terms, prompts, tool and model, date, retrieved sources, inclusion and exclusion decisions, verification and final use.
Knowledge Type: Research documentation · Normative Weight: Core
I.4-K06
Understands confidentiality, personal data, proprietary sources, licensing, copyright, legal research and the need for specialist interpretation and human review.
Knowledge Type: Protection and professional boundary · Normative Weight: Core
Applied skills · 5
I.4-S01
Define the reporting research question, required authority, jurisdiction, date, version, source hierarchy and acceptable evidence.
Skill Type: Core applied capability · Observable Output or Result: AI-assisted research plan and source hierarchy
I.4-S02
Use authorised tools and well-structured queries to identify candidate sources, terms, passages and relationships without treating outputs as authoritative.
Skill Type: AI-assisted discovery · Observable Output or Result: Candidate source and query record
I.4-S03
Open and compare candidate sources, verify issuer, version, date, scope, quotations and citations and retain the authoritative evidence.
Skill Type: Source verification · Observable Output or Result: Verified source and retrieval register
I.4-S04
Summarise verified material accurately, preserve exceptions and uncertainty and distinguish source-derived content from inference or unresolved interpretation.
Skill Type: Synthesis and limitation control · Observable Output or Result: Verified research synthesis
I.4-S05
Maintain the AI research, citation and use record and control later reuse when sources, tools or requirements change.
Skill Type: Documentation and reuse · Observable Output or Result: AI research, citation and use record
Professional behaviours · 3
I.4-B01
Does not cite an AI response, generated reference or search snippet as the authoritative basis for a material reporting conclusion.
Behaviour Type: Source authority discipline · Non-compensable Requirement: No
I.4-B02
Does not use a quotation, requirement, date or source attribution without checking it against the underlying material.
Behaviour Type: Verification diligence · Non-compensable Requirement: No
I.4-B03
Does not remove material source context or use confidential, licensed or copyrighted content outside authorised conditions.
Behaviour Type: Context and rights respect · Non-compensable Requirement: No
Typical tasks · 4
I.4-T01
Define the research question, authority, jurisdiction, date, version, source types, confidentiality and completion criteria.
Primary Output Link: I.4-O01
I.4-T02
Use authorised AI and search tools to identify candidate sources and passages and record queries, prompts, tool and date.
Primary Output Link: I.4-O03
I.4-T03
Open and verify sources, citations, quotations, tables, versions, scope and effective status and resolve conflicts or gaps.
Primary Output Link: I.4-O02
I.4-T04
Prepare a source-grounded synthesis, record limitations and retain the final research, citation and use trail.
Primary Output Link: I.4-O03
Expected outputs · 3
I.4-O01
AI-assisted research plan and source hierarchy
Output Type: Professional work product
I.4-O02
Verified source and retrieval register
Output Type: Professional work product
I.4-O03
AI research, citation and use record
Output Type: Professional work product
Proficiency indicators
Level 1 · Foundation
I.4-L1-01
Can use an approved AI research workflow, verify specified sources and maintain a research and citation record.
Indicator Dimension: Task execution
I.4-L1-02
Can identify obvious fabricated, outdated, unofficial or misquoted material and escalate uncertain technical interpretation.
Indicator Dimension: Quality, judgement and accountability
Level 2 · Practitioner
I.4-L2-01
Can independently conduct and document AI-assisted research for a moderately complex reporting question across several source types and versions.
Indicator Dimension: Task execution
I.4-L2-02
Can resolve routine source conflicts, preserve context and explain source authority, limitations and inference to management and reviewers.
Indicator Dimension: Quality, judgement and accountability
Level 3 · Advanced Practitioner
I.4-L3-01
Can design or critically review enterprise AI-assisted research and source-governance methods across standards, jurisdictions and knowledge systems.
Indicator Dimension: Method design and review
I.4-L3-02
Can challenge unverified or fabricated research, resolve significant source and version disputes and design and advise on organisation-wide AI-research rules for source verification, citation, rights and reuse, consistent with established source-authority and legal requirements.
Indicator Dimension: Leadership and governance
Illustrative evidence · 6
I.4-E01
AI-assisted research plan and source hierarchy.
Evidence Type: Work product
I.4-E02
Verified source and retrieval register with authority, version, date and location.
Evidence Type: Work product
I.4-E03
AI research, citation and use record.
Evidence Type: Work product
I.4-E04
Query, prompt, retrieval, source-opening, quotation, citation, synthesis and reuse trail.
Evidence Type: Process evidence
I.4-E05
Documented standards, legal, technical, copyright, quality or management review and the practitioner's response.
Evidence Type: Review evidence
I.4-E06
Observed verification and correction of a fabricated or context-lost AI research output.
Evidence Type: Observed performance
Assessment · 3
I.4-A-L1
AI research exercise, source-verification task and situational judgement
Research scoping; authority and version; citation and context verification; escalation.
I.4-A-L2
Integrated research case and professional memorandum
Source hierarchy; retrieval and context; verification and conflict resolution; documentation and rights.
I.4-A-L3
Complex research-governance case, portfolio and oral defence
Method design; source and version integrity; rights and reuse control; governance advice and oral defence.
Relationships · 6
FromToTypeRationale
A.2I.4Risk and control linkageAI-assisted research must operate within the source-verification and provenance requirements established here.
I.4A.2Evidence and traceability linkageAI-assisted discovery should preserve authoritative source hierarchy and current source-version control.
I.4A.5Change and lifecycle linkageAI-assisted research may support change monitoring but material amendments require verified source review.
I.4F.3Evidence and traceability linkageResearch sources and conclusions should be evaluated for relevance, reliability and sufficiency.
I.4I.6Risk and control linkageAI-retrieved or summarised information should be grounded and verified before reporting use.
I.4I.8Professional conduct and collaboration linkageResearch prompts, sources and knowledge systems should comply with confidentiality, rights and permitted-use controls.
Role profiles for this unit
RoleTarget levelRelevanceEvidence expectation
Corporate Sustainability Reporting Practitioner PractitionerRequired?A case or work sample demonstrating independent performance, documented judgement and a reviewable professional output.
Sustainability Reporting Manager or Lead PractitionerRequired?A case or work sample demonstrating independent performance, documented judgement and a reviewable professional output.
Sustainability Reporting Adviser or Consultant Advanced PractitionerRole-defining?A complex case or verified portfolio, supplemented by oral or observed defence, demonstrating method design, challenge and governance capability.
Sustainability Data, Systems and Controls Specialist PractitionerRequired?A case or work sample demonstrating independent performance, documented judgement and a reviewable professional output.
Assurance Readiness and Reporting Quality Specialist Advanced PractitionerRole-defining?A complex case or verified portfolio, supplemented by oral or observed defence, demonstrating method design, challenge and governance capability.
Investor, Capital Markets and Ratings Disclosure Specialist Advanced PractitionerRole-defining?A complex case or verified portfolio, supplemented by oral or observed defence, demonstrating method design, challenge and governance capability.