Auto-Generated AASB S2 Climate Disclosure Documents:...

Group 1 reporters built their first AASB S2 climate disclosure the same way most first-year reporters do. A sustainability lead wrote the governance and strategy sections in Word. A finance analyst assembled the metrics tables in a shared workbook. A brand designer laid out the final PDF for the annual report. Three people, three tools, six weeks of reconciliation, and one number that changed twice in the last 48 hours before sign-off.

That build-by-hand approach works under Phase 1 limited assurance. It stops working under reasonable assurance.

The reason is straightforward. Under ASSA 5010 reasonable assurance, the auditor walks every disclosed metric back to source. Every Scope 1 diesel litre. Every Scope 2 kWh. Every methodology label. Every prior-year comparative. The workbook cell that survived limited-assurance sampling gets tested end-to-end. If the number in the PDF does not reconcile to the ledger, or the ledger does not reconcile to the source document, the assurance opinion is not clean.

So the question we kept running into when we designed the AASB S2 disclosure output was this. What is safe to auto-generate, and what deliberately stays in the reporter's hands?

The AASB S2 disclosure structure and where it breaks

AASB S2 organises the climate-related financial disclosure into four pillars. Governance. Strategy. Risk management. Metrics and targets. The AASB has published this in AASB S2 paragraphs 6 through 37 and the associated application guidance.

The narrative pillars (governance, strategy, risk management) are author-driven. A board-oversight statement, a scenario analysis discussion, a transition plan narrative. These come from the reporter's own governance record and cannot be generated from a data platform.

The metrics and targets pillar is different. Paragraph 29 lists what has to be disclosed. Gross Scope 1 emissions. Gross Scope 2 (both location-based and market-based where material). Gross Scope 3 by category. GWP basis. Measurement approach. Methodology disaggregation. Prior-year comparatives under paragraph 21. Industry-based metrics under paragraph 32. And any target the reporter has set under paragraphs 33 to 36.

That is the section a data platform can populate from the ledger. And it is the section where reasonable assurance concentrates its walk-through.

If you want the strategic framing on where automation ends and authorship begins, we covered it in how to generate an AASB S2 climate disclosure report. This post is about the mechanics of the document that comes out the other side.

What auto-generation covers

The disclosure output is built from the emission ledger. Not from a spreadsheet copy of the ledger, not from a static export, but from the live records the platform holds. So Scope 1, Scope 2, and all fifteen Scope 3 categories are pulled straight from the underlying data.

A few specifics matter here.

Methodology labels sit on every record. Activity-based, hybrid, supplier-specific, average-data, spend-based. Paragraph 29(a)(vi) requires disaggregation by measurement approach. If the labels only exist at the reporting layer, the auditor asks for the record-level breakdown and the process stalls. The label belongs on the emission record itself, so the paragraph 29(a)(vi) table is a data query, not a manual reconciliation.

Prior-year comparatives should come from the locked prior period. Paragraph 21 requires comparative information for the previous reporting period. Reporters who tried to recompute last year's numbers using this year's factor library in year one learned the hard way that the comparative moves and the auditor asks why. Period locking preserves the numbers as they were disclosed, with a recorded reason for any restatement, so the comparative in this year's document matches the comparative in last year's document.

The GWP basis has to be stated. Carbonly writes AR5 values, the set the Clean Energy Regulator requires for NGER. Where the AASB S2 disclosure needs AR6 values, that recalculation sits outside the platform and should be documented alongside the disclosure. We covered the mechanics of that split in AR5 vs AR6 methane reporting for NGER and AASB S2.

Gross versus net disclosure is separated. If the reporter uses avoided emissions or carbon offsets in a public net-zero claim, the disclosure output shows gross emissions and the net position as two distinct numbers. AASB S2 paragraph 29 is a gross-emissions standard. The net position is a supplementary disclosure and is presented as one.

Energy consumption disclosures come through the same ledger. Total energy in gigajoules, split by renewable and non-renewable sources where the source data supports it.

Natural resource metrics

AASB S2 pulls in industry-based metrics through paragraph 32 and the associated Application Guidance §B40 to §B44 and §B56. For reporters in mining, oil and gas, agriculture, and water-intensive processing, the disclosure has to carry natural-resource metrics alongside the emissions numbers, not folded into them.

One way to handle this is to hold water consumption (in kilolitres) and recycled material throughput (in tonnes) as reporting-only records with a zero emission factor, so the Scope 1/2/3 CO2e maths is untouched but the physical quantity can flow into the disclosure document.

For a water utility, a food and beverage manufacturer, or a mining operation, this is the difference between a disclosure that is defensible and a disclosure that misses a paragraph 32 requirement the auditor was going to ask about anyway.

Paragraph mapping

Every metrics section in the generated document carries the AASB S2 paragraph reference in the section header. Paragraph 21 for comparatives. Paragraph 29(a) for gross Scope 1, 2, and 3. Paragraph 29(a)(vi) for the methodology disaggregation. Paragraph 29(g) for the GWP basis. Paragraph 32 for industry metrics. Paragraphs 33 to 36 for targets.

The purpose is not decorative. Reasonable assurance under ASSA 5010 walks the disclosure against the standard. When the paragraph reference sits on the section header, the auditor's walk-through is a one-page cross-reference. When the reference is missing, the auditor asks the reporter to hand-map the disclosure to the standard, and the engagement clock keeps running.

Paragraph 22 (scenario analysis) is a placeholder in the auto-generated document. The scenario analysis is not something the platform can generate. It is a strategic exercise the reporter runs against 1.5, 2 and 3-degree pathways and the platform reserves space for the narrative to land. Climate scenario analysis for AASB S2 walks through what belongs in that section.

Format outputs: PDF and Word

The generator produces both a PDF and a Word document from the same underlying disclosure payload.

The PDF is the final artefact. It is what most reporters attach to the annual report or publish as a standalone sustainability report. It carries the paragraph references, the metrics tables, the natural resource metrics, the target progress against baseline, and the inline charts.

The Word document is the working artefact. The metrics sections are pre-populated identically to the PDF. The narrative sections carry structured placeholders (H2 headings for governance, strategy, risk management, plus the paragraph 22 scenario section) with a short brief inside each explaining what belongs. The internal author writes the narrative around the auto-generated numbers without having to reformat tables or rebuild charts.

Both outputs are generated from the same disclosure payload, so the metrics in the PDF and the metrics in the Word come from the same numbers. This is a small point but a load-bearing one. Year one reporters commonly had a PDF version and a Word version that drifted by one or two decimal places, and the drift was almost always the last thing anyone caught before print.

Charts and visualisations

Charts are rendered inline. A bar chart of Scope 1, 2 and 3 by category. A stacked area of emissions by facility over the reporting period. A trajectory chart against baseline for reporters covered by the Safeguard Mechanism. A comparison of prior-year and current-year for the paragraph 21 comparative section.

Chart type is chosen per metric. A pie chart is not used for scope breakdowns because the auditor and the audit committee find the numeric comparison harder to read than a bar. A line chart is used for trajectory. A stacked area is used for facility-level composition where the total also matters.

Each chart is embedded in both the PDF and the Word output at native resolution. The reporter can replace a chart in the Word file with a designer's version for the annual report layout without breaking the reconciliation because the underlying numbers are still in the table alongside.

Executive Summary Generator integration

The same emission ledger feeds the Executive Summary Generator. That is what produces the AGM slide, the board-pack chart set, and the short-form investor summary.

The reason this matters is reconciliation. When the AASB S2 disclosure says total Scope 1 was 84,230 tonnes CO2-e and the AGM slide says total Scope 1 was 84,180 tonnes CO2-e, someone in the audience will notice and someone on the audit committee will ask why. Both artefacts pull from the same source, so the numbers are the same by construction.

This is the same argument for the ASX200 carbon compliance workflow and for the CFO who has to sign the disclosure. One ledger. Multiple outputs. No parallel spreadsheet.

Audit trail on the document itself

Carbonly records typed audit events on ledger writes. Not log lines. Structured events that show who changed what in the ledger, and when.

Under ASSA 5010, the assurance provider may ask for evidence that the disclosure attached to the annual report came from the same ledger state as the disclosure the audit committee signed off. Keep the generated file and the signed-off version together, with the period locked, so that question is easy to answer.

The evidence pack for assurance links each emission record back to its source document. We walked through the Evidence Pack workflow in NGER audit preparation with an evidence pack and the assurance mechanics in reasonable assurance ASSA 5010 readiness.

The restatement path

Prior-year disclosures get restated more often than reporters expect. A supplier updates their factor. A new activity type is discovered mid-year. A methodology change flips a Scope 3 category from spend-based to activity-based.

When a locked period is restated, Carbonly records the restatement reason, so the written justification travels with the restated figures and a future auditor can see why the number moved.

This is the mechanic AASB S2 paragraph 21 anticipates. Restatement is not an admission of failure. It is a controlled process, and the disclosure output has to show the trail. Restating by editing last year's PDF is risky, because an auditor will usually want the version history and the justification, not a new PDF.

What the disclosure output does not do

We are upfront about this because a disclosure tool that promises more than it delivers is a liability.

The output does not write the governance narrative. Who sits on the climate risk committee, how often it meets, what the escalation path to the board is. That is authorship.

The output does not write the strategy discussion. The reporter's positioning against the transition to a low-carbon economy, the resilience of the business model, the transition plan itself. The transition plan template is a separate exercise.

The output does not write the scenario analysis. The 1.5, 2 and 3-degree pathway modelling and the associated financial impact discussion. Paragraph 22 requires it and paragraph B12 explains the granularity. The platform reserves the placeholder and the narrative fills it.

The output does not write the risk management response. The identification, assessment and prioritisation of climate risks and opportunities. That work is captured in the climate risk register and the reporter maps it into the disclosure.

We do not have a built-in PCAF data quality score, a GRESB submission generator, a NABERS rating output, a CBAM template, a CSRD or ESRS export, or a direct integration with Climate Active or the ISCA IS rating tool. We hold the underlying data. Those framework views sit on top of the ledger and stay in the reporter's authorship for now.

The consultant angle

A sustainability consultant running an AASB S2 engagement for a client can generate the disclosure output early. Metrics section, natural resource metrics, paragraph references, prior-year comparatives, chart set. The metrics are defensible from day one because they come from the client's ledger and every record ties back to a source document.

The consultant then does the work only the consultant can do. Governance framing. Strategy narrative. Scenario analysis. Risk management discussion. Transition plan. The final document goes to the client with the ledger intact for the client's ongoing use. Next year, the client renews the disclosure without a fresh six-week rebuild.

This is the workflow we designed the AASB S2 disclosure output for. The consultant is the craftsperson. Carbonly is the workshop equipment.

What to do first

Two things.

Run a dry generation against a completed quarter. Not the full year. A quarter is enough to see the paragraph references, the methodology disaggregation, the natural resource metrics, and the chart output. If a Scope 3 category is missing a methodology label, you want to find out at a dry run, not at Phase 2 assurance.

Then walk the output past the assurance provider before the reporting period ends. Not for sign-off, and not as a preview of the final report. As an alignment on paragraph mapping, methodology disaggregation, and the evidence trail. Auditors will tell you whether the trail meets their sampling protocol if you ask early. If you ask late, the answer costs money.

Carbonly is priced per project with a $100 per month workspace minimum. Reach out at hello@carbonly.ai for a walkthrough of the AASB S2 disclosure output on your own ledger.

Auto-Generated AASB S2 Climate Disclosure Documents: What Ships in the PDF and Word Output