The European Union has scaled back the scope of mandatory ESG reporting, but has not resolved the problem that the first few years of the CSRD have highlighted so clearly. In many companies, non-financial data still operates outside the corporate management system: it is scattered across ERP, HR, procurement, technical systems, supplier invoices and spreadsheets. The report merely highlights the consequences of this fragmentation.
Following the changes introduced by the Omnibus I Directive, mandatory CSRD reporting essentially covers companies with both more than 1,000 employees and net revenue exceeding €450 million. On 3 July 2026, the European Commission also adopted the simplified ESRS. The number of mandatory data points is set to fall by over 60 per cent, and the total number by over 70 per cent. The Commission estimates that this could reduce reporting costs by over 30 per cent per company. The standards are still awaiting final approval by the European Parliament and the Council.
Regulatory burdens are therefore lighter. Operationally, however, this changes little for a company that is unable to quickly answer questions such as how much energy a specific plant consumes, how emissions in the supply chain were calculated, or which system the figure published in the report comes from.
The problem begins with finance
The first wave of reporting revealed a significant gap between ESG narratives and financial information. Deloitte analysed 126 ESRS reports from nine Central European countries. Only 14 per cent of companies quantified the financial impact of sustainability-related risks and opportunities. 39 per cent limited themselves to a qualitative description.
This is a fundamental difference. Information on climate risk has limited managerial value until it can be linked to the cost of assets, CapEx, Opex, margins, insurance or cash flow. At this point, ESG ceases to be a report and becomes a matter of economic accounting.
At the same time, the quality of the process still falls short of the standard seen in finance. In a KPMG survey of 1,320 board members and managers, 76 per cent of companies remained at an early or intermediate stage of maturity in terms of ESG reporting and data assurance.
Data processing remains the weakest link
A 2025 PwC survey, conducted amongst 496 companies from 40 countries, illustrates the scale of the problem more directly. Among organisations preparing at that time for the next wave of CSRD, as many as 88 per cent were using spreadsheets to collect sustainability data. Companies that had already completed their reporting most frequently cited the need for earlier checks on data completeness, better use of technology and stronger cross-functional collaboration. 37 per cent also highlighted the need to involve the auditor at an earlier stage.
Excel is not the problem in itself. The problem arises when it becomes an integration layer between dozens of systems, departments and data owners. In such cases, every subsequent update requires manual reconciliation, file modifications, version control and the reconstruction of source data for audit purposes.
This is well illustrated by the case of the Australian Gas Infrastructure Group. The company had over 35 ESG data sources spread across three business units. After building a shared data platform on Microsoft Fabric, it automated their integration. According to a case study published by Fujitsu, the sustainability team saved the equivalent of five to six weeks’ work, and the preparation of one of the KPI reporting processes – which previously took two days and involved four teams – was automated.
This better illustrates the economics of ESG than the sheer number of required indicators. A cost is incurred every time data is manually retrieved, recalculated, reconciled and approved.
Technology won’t fix poor architecture
Companies are increasingly trying to streamline this process using AI. According to a PwC survey, the use of artificial intelligence in sustainability reporting rose from 11 per cent to 28 per cent over the course of a year. AI is used, amongst other things, to identify risks and to collect, integrate and validate data. At the same time, most implementations remain at the experimental and pilot stage.
The reason is straightforward: whilst a model can analyse data more quickly, it cannot determine the organisation to which the data belongs, its definitions or the source of truth. If emissions, employment figures or energy consumption are calculated differently across different companies, automation will merely consolidate these inconsistencies more quickly.
Therefore, following the simplification of the CSRD, the most significant change does not concern the number of pages in the report. It concerns the criterion of investment viability. A system built solely for annual ESG disclosure is a compliance cost. Infrastructure that links the same data to finance, procurement, production and risk management is already part of the management system.
And it is precisely at this juncture that it is determined whether ESG is a financial, operational or technological issue for a company. Most often, it is a problem with the quality of the entire data model.
