Domain I · Area I-CA2 · Responsible AI use, grounding and model-risk control

I.5Use AI for mapping, analysis and controlled drafting

Working Draft version 1.1

Purpose and scope

This unit covers the capability to use AI to assist standards and disclosure mapping, classification, gap analysis, extraction, comparison, data or narrative analysis, scenario preparation, workflow design and controlled drafting. It includes task decomposition, prompt and context design, approved inputs, structured outputs, human review, materiality and judgement limits, versioning and integration into controlled reporting workpapers and decisions.

Applied competency statement

Can use AI proportionately to accelerate mapping, classification, analysis and drafting while preserving approved inputs, professional judgement, human review, source and evidence links, version control and reporting accountability.

Boundary and escalation

This unit owns controlled AI assistance for mapping, classification, analysis and drafting. H.3 owns significant professional judgement, I.6 owns grounding and verification and I.7 owns use-case and model risk. Final materiality, legal, accounting, governance, assurance, ethical, sign-off and publication decisions remain with authorised people.

Key quality risks

Using AI before the task, criteria or authoritative inputs are defined; confidential or irrelevant context supplied; prompt wording designed to obtain a preferred conclusion; standards text reproduced without rights or verification; classifications accepted without edge-case review; calculations or tables generated without reperformance; boilerplate drafting; materiality, legal, governance, assurance or publication decisions delegated to AI; model output copied into final reporting; and prompt, model, source, review or revision history not retained.

Required knowledge · 6

I.5-K01

Understands suitable uses such as mapping, extraction, classification, comparison, gap analysis, summarisation, analytical support, template generation and controlled drafting and unsuitable or high-risk uses.

Knowledge Type: AI-assisted reporting use cases · Normative Weight: Core

I.5-K02

Understands task decomposition, instructions, criteria, examples, constraints, output schema, context, iteration, prompt injection and the risks of leading prompts.

Knowledge Type: Task and prompt design · Normative Weight: Core

I.5-K03

Understands authoritative inputs, source and version control, completeness, relevance, context-window limitations, chunking, confidential data and the effect of poor inputs.

Knowledge Type: Input and context quality · Normative Weight: Core

I.5-K04

Understands structured outputs, identifiers, evidence references, uncertainty fields, alternatives, confidence indicators where appropriate and integration into workpapers and workflows.

Knowledge Type: Output structure and integration · Normative Weight: Core

I.5-K05

Understands review depth, re-performance, sampling, materiality, exception and specialist escalation and the non-delegable nature of significant reporting decisions.

Knowledge Type: Human review and judgement · Normative Weight: Core

I.5-K06

Understands use-case approval, prompt, model and version, inputs, outputs, edits, reviewer, final decision, rights and the distinction between AI assistance and professional authorship or approval.

Knowledge Type: Records and boundary · Normative Weight: Core

Applied skills · 5

I.5-S01

Select and define an AI use case, task, criteria, approved inputs, decision boundaries, reviewer and completion conditions.

Skill Type: Core applied capability · Observable Output or Result: AI use-case, prompt and context specification

I.5-S02

Develop clear, neutral prompts, examples, constraints and structured output formats and provide only approved, relevant and protected context.

Skill Type: Prompt and context design · Observable Output or Result: Controlled prompt and context package

I.5-S03

Generate and iterate mapping, classification, analysis or drafting outputs while retaining sources, versions, alternatives and uncertainty.

Skill Type: AI-assisted execution · Observable Output or Result: AI-assisted mapping, analysis or drafting workpaper

I.5-S04

Verify evidence, reperform material calculations or classifications, challenge omissions and bias and revise the output into a professional work product.

Skill Type: Human review and revision · Observable Output or Result: Human review, revision and approval record

I.5-S05

Link the approved result to requirements, decisions, evidence and workflows and reassess it when tools, sources or reporting circumstances change.

Skill Type: Integration and change control · Observable Output or Result: AI-assisted work integration and change record

Professional behaviours · 3

I.5-B01

Does not delegate significant materiality, legal, governance, assurance, ethical or publication judgement to an AI system.

Behaviour Type: Human judgement preservation · Non-compensable Requirement: No

I.5-B02

Does not design prompts or select outputs to manufacture a preferred reporting conclusion and retains material instructions and revisions.

Behaviour Type: Prompt neutrality and transparency · Non-compensable Requirement: No

I.5-B03

Does not copy AI output into reporting as if it were verified professional work without substantive human review, correction and approval.

Behaviour Type: Professional authorship · Non-compensable Requirement: No

Typical tasks · 4

I.5-T01

Define the AI use case, reporting task, criteria, approved context, prohibited decisions, output structure, evidence and reviewer.

Primary Output Link: I.5-O01

I.5-T02

Prepare prompts and context and generate structured candidate outputs with source, uncertainty and exception fields.

Primary Output Link: I.5-O02

I.5-T03

Verify and reperform material elements, review edge cases, omissions, bias and language and revise the result.

Primary Output Link: I.5-O03

I.5-T04

Integrate the approved output into workpapers and workflows and control tool, prompt, source, model and later change.

Primary Output Link: I.5-O03

Expected outputs · 3

I.5-O01

AI use-case, prompt and context specification

Output Type: Professional work product

I.5-O02

AI-assisted mapping, analysis or drafting workpaper

Output Type: Professional work product

I.5-O03

Human review, revision and approval record

Output Type: Professional work product

Proficiency indicators

Level 1 · Foundation

I.5-L1-01

Can use an approved prompt and context package, generate a defined reporting output and complete the required human review checklist.

Indicator Dimension: Task execution

I.5-L1-02

Can identify obvious unsupported, boilerplate, omitted or judgement-sensitive AI output and stop or escalate it.

Indicator Dimension: Quality, judgement and accountability

Level 2 · Practitioner

I.5-L2-01

Can independently design and execute a moderately complex AI-assisted mapping, analysis or drafting use case and convert the output into a controlled professional work product.

Indicator Dimension: Task execution

I.5-L2-02

Can determine review depth, resolve routine errors and bias and explain tool use, limitations and human decisions to management and reviewers.

Indicator Dimension: Quality, judgement and accountability

Level 3 · Advanced Practitioner

I.5-L3-01

Can design or critically review enterprise AI-assisted reporting methods, prompt and context standards and human-review controls across complex use cases.

Indicator Dimension: Method design and review

I.5-L3-02

Can challenge unsafe delegation or unverified output, resolve significant AI-assisted judgement and integration failures and advise governance bodies on use, restriction and residual risk.

Indicator Dimension: Leadership and governance

Illustrative evidence · 6

I.5-E01

AI use-case, prompt and context specification.

Evidence Type: Work product

I.5-E02

AI-assisted mapping, analysis or drafting workpaper.

Evidence Type: Work product

I.5-E03

Human review, revision and approval record.

Evidence Type: Work product

I.5-E04

Use-case, prompt, context, model, source, output, revision, approval and integration trail.

Evidence Type: Process evidence

I.5-E05

Documented technical, legal, data, specialist, quality or management review and the practitioner's response.

Evidence Type: Review evidence

I.5-E06

Observed use and correction of AI in a mapping, analytical or drafting task.

Evidence Type: Observed performance

Assessment · 3

I.5-A-L1

Controlled AI-use exercise, output review and situational judgement

Use-case and instruction compliance; source and output review; human judgement and escalation.

I.5-A-L2

Integrated AI-assisted reporting case and professional memorandum

Use-case and prompt design; source and context control; substantive review; professional integration and documentation.

I.5-A-L3

Complex AI-use governance case, portfolio and oral defence

Method design; non-delegable judgement; prompt, source and review controls; governance advice and oral defence.

Relationships · 6

From	To	Type	Rationale
I.5	G.5	Disclosure, claims and publication linkage	AI may assist disclosure mapping and outlines but should follow approved architecture, criteria and evidence.
I.5	G.6	Disclosure, claims and publication linkage	AI-assisted drafting requires entity-specific human revision, balance and source traceability.
I.5	H.3	Professional conduct and collaboration linkage	Significant reporting judgements remain human and should not be delegated to AI.
I.5	I.6	Risk and control linkage	AI-assisted mapping, analysis and drafting require source, calculation and context grounding.
I.5	I.7	Risk and control linkage	Use-case, model and bias risk determine restrictions, testing and review depth.
I.6	I.5	Operational and workflow linkage	Verification converts candidate AI-assisted work into a controlled professional work product.

Role profiles for this unit

Role	Target level	Relevance	Evidence expectation
Corporate Sustainability Reporting Practitioner	Practitioner	Required?	A case or work sample demonstrating independent performance, documented judgement and a reviewable professional output.
Sustainability Reporting Manager or Lead	Advanced Practitioner	Role-defining?	A complex case or verified portfolio, supplemented by oral or observed defence, demonstrating method design, challenge and governance capability.
Sustainability Reporting Adviser or Consultant	Advanced Practitioner	Role-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	Advanced Practitioner	Role-defining?	A complex case or verified portfolio, supplemented by oral or observed defence, demonstrating method design, challenge and governance capability.
Assurance Readiness and Reporting Quality Specialist	Advanced Practitioner	Role-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 Practitioner	Role-defining?	A complex case or verified portfolio, supplemented by oral or observed defence, demonstrating method design, challenge and governance capability.