The Real State of Carbon Accounting in Australia: A Practitioner's View

We've spent the last eighteen months talking to sustainability managers, CFOs, auditors, and climate consultants across Australia. The conversations follow a pattern. Someone describes what they thought mandatory climate reporting would involve. Then they describe what it actually involves. The gap between those two things is where most of the pain lives right now.

So here's our honest assessment of carbon accounting in Australia in 2026. Not the version you'd read in a Big Four thought leadership piece. Not the vendor pitch. The version you'd hear over a coffee at a sustainability conference after the official sessions are done — when people stop performing confidence and start admitting what's hard.

Group 1 reports landed. They weren't pretty.

The first mandatory sustainability reports under AASB S2 started hitting ASIC's desk from March 2026. Group 1 entities — the ASX 200 and their private equivalents, roughly companies with $500M+ revenue, $1B+ gross assets, or 500+ employees — had been preparing since January 2025. Eighteen months of lead time. Big budgets. Big Four advisory teams.

And yet ASIC's early review of voluntary climate disclosures (which many Group 1 entities had been producing before the mandate) found disclosures that were "often repetitive, with key information about the management of climate-related risks and opportunities missing or unclear." Scenario analysis sections lacked detail on underlying assumptions. Transition plans weren't linked to actual targets or strategies.

That was the voluntary stuff. The bar for mandatory reports under the Corporations Act is higher.

Oxford Economics flagged three recurring problems from the first wave of reporters that we've heard echoed by almost every auditor we've spoken to this year. First, companies treated year one as a one-off project — they paid consultants $150K to $400K to build a disclosure model that couldn't be repeated in year two without the same consultants. Second, scenario analysis used global pathways (IEA Net Zero 2050, NGFS scenarios) without calibrating them to Australian conditions, specific revenue lines, or asset valuations. Third — and this is the most damaging one — financial impact estimates changed materially once assurance providers started testing them. If your numbers move when someone checks your work, you have a methodology problem.

The quality gap is real

Here's a number that should worry people. The Australian Institute of Company Directors found that only 45% of surveyed directors believe their board has adequate skills and experience to address climate issues. That's down from where it was three years ago. Director confidence in their own ability to oversee climate risk is declining at the exact moment those risks are becoming legally material.

And this isn't just a governance problem. It flows through to the numbers. The ANAO's performance audit of NGER found that 72% of 545 reports examined contained errors. Seventeen percent had significant errors. That audit covered earlier NGER reporting years, but the underlying cause — manual data collection, inconsistent methodologies, and insufficient quality controls — hasn't gone away. If anything, ASRS makes it worse because you're now doing the same work under director liability provisions that carry penalties up to $15 million or 10% of annual turnover.

We keep coming back to a simple observation: the quality of emissions data in Australia is not where it needs to be. Not close. And no amount of glossy reporting will fix bad input data.

The consultant bottleneck nobody planned for

Australia doesn't have enough qualified people to do this work. That's not an opinion — it's arithmetic.

The EY Corporate Reporting Survey found that only 29% of finance leaders believe their teams have the skills to manage sustainability-related reporting. Very few Australian accounting or finance degree programs teach ISSB-aligned climate disclosure, scenario analysis, Scope 3 accounting, or ESG assurance. The World Economic Forum estimates that only one in eight workers globally have sustainability skills, and Australia's position is no better.

So everyone's hiring the same consultants. And those consultants are at capacity.

We've heard from multiple mid-market companies that Big Four engagement timelines for ASRS advisory work have blown out to 6-8 months. Some couldn't get a team at all for the 2025-26 financial year. Mid-tier firms like BDO, Grant Thornton, and PKF have picked up overflow work, but they're facing their own capacity constraints. The result is a two-speed market: large companies who locked in advisory teams early and got their disclosures done (with varying quality), and everyone else who's scrambling.

This bottleneck won't ease with Group 2 reporting. It'll get worse. Group 2 entities — $200M+ revenue, $500M+ gross assets, or 250+ employees, plus all NGER reporters — start reporting for financial years from 1 July 2026. That pulls in hundreds of additional companies, many of whom have never done formal emissions reporting. They all need consultants, assurance providers, and software. At the same time.

The management consulting market in Australia is projected to grow from $9.4 billion in 2026 to $12.7 billion by 2031, and mandatory sustainability reporting is one of the primary drivers. That growth is a symptom of the problem, not a solution to it.

What Group 2 is actually facing

Let's be specific about what Group 2 entities need to produce, because we still talk to companies who think this is a "sustainability report" in the old sense — a PDF with some recycling stats and a CEO letter about values.

It's not. It's a statutory financial disclosure. It goes inside your annual report. Directors sign off on it. An assurance provider reviews it. ASIC can (and will) scrutinise it. The same liability framework that applies to your profit and loss statement applies to your climate disclosures.

Year one requirements under AASB S2 include governance disclosures (evidence that your board actually oversees climate risk — not a committee that never meets), strategy disclosures with scenario analysis, risk management process disclosures, and Scope 1 and 2 emissions with limited assurance. Scope 3 gets a one-year deferral. But only one year.

Scope 1 and 2 emissions have no safe harbour protection. Full stop. If those numbers are wrong, you're exposed from day one. The modified liability regime that protects forward-looking statements and Scope 3 data doesn't extend to your direct and energy emissions. You need those figures right — calculated using the GHG Protocol, sourced from auditable data, with the correct NGA emission factors for your state and activity.

We built Carbonly around this exact problem because we watched companies spend weeks manually extracting consumption data from utility bills — the raw material for Scope 1 and 2 calculations — when that process can be automated down to minutes. But automation alone doesn't fix everything. You still need someone who understands the methodology, can defend the numbers to an assurance provider, and knows whether to use AR5 or AR6 Global Warming Potential values (NGER uses AR5; AASB S2 requires AR6 — a technical difference that creates real headaches for dual reporters).

The assurance crunch

ASSA 5010, released by the AUASB in January 2025, sets out Australia's sustainability assurance framework. It's deliberately staged — limited assurance in years one through three, then reasonable assurance from financial years beginning 1 July 2030.

That staging exists for a reason. Australia doesn't have enough qualified sustainability assurance practitioners to provide reasonable assurance across thousands of entities right now. The standard explicitly acknowledges "the capacity and capabilities of auditors and their experts during the initial years of reporting."

Even limited assurance is stretching capacity. A mid-market company can expect to pay $30,000 to $80,000 for limited assurance over their first ASRS disclosure. That's before the assurance provider asks you to fix the data quality issues they find during the engagement — which, based on what we've seen from Group 1, happens in the majority of cases.

When reasonable assurance kicks in from 2030, the bar rises significantly. Reasonable assurance means the auditor needs to positively confirm your numbers are materially correct, not just that nothing came to their attention suggesting otherwise. The infrastructure to support that level of assurance across every ASRS reporter in Australia doesn't exist yet. We're not sure it'll exist by 2030 either.

ASIC has signalled it will take "a pragmatic and proportionate approach to supervision and enforcement as industry adjusts." But pragmatic has limits. They've already imposed $34.7 million in greenwashing penalties across three cases in 2024-25 alone — Mercer Super ($11.3M), Vanguard ($12.9M), and Active Super ($10.5M). The ACCC has added another $8.25 million through the Clorox/GLAD bags case. When regulators say pragmatic, they mean "we won't punish honest effort." They don't mean "we'll look the other way."

Scope 3: the elephant that's about to sit on everyone

Every practitioner we know has the same private assessment of Scope 3 emissions data: it's a mess.

That's not a controversial opinion. Scope 3 represents roughly 90% of most organisations' total carbon footprint, and the data quality underpinning those estimates is — to put it diplomatically — early-stage. Most companies are using spend-based estimates with Environmentally Extended Input-Output (EEIO) factors, which respond to changes in spending but not to actual emission reductions in the supply chain. You could switch to a supplier that's halved their emissions, and your Scope 3 number wouldn't change because you're still spending the same amount.

Group 1 entities need Scope 3 in their second reporting period. For those with June year-ends, that's the financial year ending 30 June 2027. Group 2 gets the same one-year deferral, pushing their Scope 3 deadline to the 2027-28 financial year. But the data collection needs to start well before the reporting deadline. If you haven't begun supplier engagement for Scope 3 by now, you're already behind.

Here's what makes this genuinely hard. Your Scope 3 is your supplier's Scope 1 and 2. But many of those suppliers — particularly SMEs — don't measure their own emissions. They don't have the systems, the knowledge, or the motivation. A 2023 study in npj Climate Action found that supply-chain data sharing remains one of the most significant barriers to accurate Scope 3 reporting, and the problem is compounded in Australia by the geographic spread of supply chains and the prevalence of small businesses in key sectors like agriculture, construction, and services.

We don't have a clean answer for this yet. Nobody does. The AASB's December 2025 amendments to AASB S2 (AASB S2025-1) introduced some targeted relief — including clarifications on Scope 3 Category 15 financed emissions and flexibility on measurement approaches — but the fundamental data quality problem persists. The modified liability provisions protect Scope 3 disclosures in the early years (only ASIC can act, and only through injunctions or declarations). That's a legal safety net, not a quality standard.

Our honest view: Scope 3 reporting in Australia will be directionally useful but quantitatively unreliable for at least the next three to four years. Companies that invest in building primary data collection systems now — instead of relying entirely on spend-based proxies — will be in a vastly better position by 2029 when assurance requirements tighten.

The software market is maturing. Slowly.

The Australian and New Zealand carbon accounting software market was valued at USD $150.6 million in 2023 and is projected to reach $668 million by 2031, growing at a 20.5% CAGR. That growth reflects mandatory reporting demand, not voluntary enthusiasm.

But here's what we see from inside the market that the market research reports miss: most companies still aren't using dedicated carbon accounting software. They're using Excel. Or they've bought software and are still using Excel alongside it because the software doesn't handle their actual data ingestion problem.

The data ingestion problem is this: carbon accounting requires consumption data (kWh, GJ, litres, tonnes) extracted from source documents — utility bills, fuel receipts, waste manifests, fleet logs. Most carbon accounting software in Australia expects you to type that data in manually or upload it in a clean CSV. The extraction step — getting numbers out of a scanned AGL bill or an Origin gas invoice where the format changes every quarter — is left to the user. That's still where most of the manual effort sits.

We built Carbonly's seven-phase AI pipeline specifically to solve this because our team spent years watching exactly this problem at BHP, Rio Tinto, and Senex Energy. At scale — hundreds of sites, thousands of bills — manual extraction is not just inefficient. It's a data quality risk. The JAMIA study found a 3.7% error rate in manual data transcription. Apply that across 500 utility bills feeding into your NGER and ASRS reports, and you've got roughly 18 entries with potential errors, each one a compliance exposure.

The market is splitting into tiers. Enterprise platforms (IBM Envizi, Persefoni, Avarni) serve the big end of town. SME tools (Trace, NetNada, Sumday) target smaller businesses with simpler needs. But there's a genuine gap in the mid-market — companies with 100 to 500 employees, 20 to 200 sites, mandatory reporting obligations, and budgets that don't stretch to $100K+ per year in software licensing. That's where we're focused, and where we think the biggest unmet need sits.

What's actually working

Not everything is broken. Some things are going well that deserve acknowledgement.

The regulatory framework itself is well-designed. ASRS aligns with the ISSB standards (IFRS S1 and S2), which means Australian companies aren't building a bespoke reporting system — they're implementing a global framework with local calibration. The NGA Factors, published annually by DCCEEW, give Australian reporters a credible, government-maintained emission factor set. State-based grid factors — NSW at 0.64 kg CO2-e/kWh, Victoria at 0.78, Tasmania at 0.20 — reflect actual generation mix differences and are updated regularly.

The phased implementation is working better than the EU's approach with CSRD, which tried to do too much too fast. Australia's three-group rollout, with Scope 3 deferrals and staged assurance requirements, gives companies time to build capability. Not enough time — but some.

NGER, for all its flaws, gave Australia a head start. The 961 registered NGER reporters already have emissions measurement systems, even if many of those systems are clunky. They have institutional knowledge of emission factors, facility boundaries, and operational control tests. When those NGER reporters get pulled into ASRS Group 2 — which they do automatically — they're not starting from zero.

And director attention is real. Climate risk has moved from a sustainability team concern to a board agenda item. It had to — the personal liability provisions in the Corporations Act made sure of that. But the effect is genuine. Boards are asking harder questions. Capital allocation is starting to factor in transition risk. It's slow, but it's happening.

Where this goes from here

We'll make three predictions about carbon accounting in Australia over the next three years. We might be wrong on the specifics, but we're confident about the direction.

First, automation will eat manual reporting. The consultant-led, spreadsheet-based model of carbon accounting is not sustainable (no pun intended) when thousands of entities report annually under assurance. Software that can extract data from source documents, apply the correct emission factors, maintain audit trails, and produce framework-aligned outputs will replace manual processes — not because it's better in theory, but because there literally aren't enough people to do it the old way. The 20.5% CAGR in the Australian carbon accounting software market reflects this inevitability.

Second, Scope 3 data quality will improve unevenly. Large companies will push primary data requirements down their supply chains. Some suppliers will respond. Many won't. Industry-level averages and benchmarks will get better as more company-level data feeds into the system. But we're still three to five years from Scope 3 numbers that anyone would describe as reliable at a portfolio level. The companies that build actual supplier engagement programs — not just annual email blasts — will have materially better data than those that don't.

Third, the gap between compliance and strategy will close. Right now, most companies treat ASRS as a compliance burden — something to survive, not use. But the companies whose climate disclosures actually inform capital allocation decisions, operational changes, and risk management will outperform. The data these disclosures require — energy consumption by site, emission intensity by product line, climate risk exposure by geography — is the same data you need to make good business decisions. The companies that figure this out first won't just report better. They'll operate better.

We don't know exactly how fast any of this will happen. We're building Carbonly on the assumption that speed matters — that companies who can go from source documents to assured emissions data in hours instead of weeks will have a real advantage. But we're also honest that technology alone won't fix the skills gap, the assurance crunch, or the Scope 3 data quality problem. Those need time, investment, and institutional learning.

What we do know is this: carbon accounting in Australia in 2026 is messy, expensive, under-resourced, and mandatory. The companies that accept all four of those realities and build systems accordingly — instead of pretending any one of them will magically resolve itself — are the ones we want to work with.

Related Reading: