GRESB retired two indicators from its 2026 Standard and told participants the point was to reduce their reporting burden. It also began scoring embodied carbon, and asking for documents to prove it.
Both were announced in the same release, and the combination is the more interesting story. A benchmark can make filing easier or it can make scoring harder. Doing both at once is a deliberate act of calibration, and it’s the kind of decision that sits with a board rather than a methodology team. Dame Alison Rose took the chair of GRESB B.V. in July 2025; the 2026 Standards Updates were released that October.
What changed in the 2026 GRESB Standard?
Four things, and none of them are cosmetic.
LE3 and SE2.2 are gone. The sustainability objectives and employee engagement indicators were retired outright. Embodied carbon moves into scored territory: TC3 and TC4 rise from 1.5 to 2.5 points each on the performance side, and DMA2 and DWS1 bring new measurement and disclosure scoring to developments that previously went unscored. Emissions from landlord-controlled tenant spaces shift from Scope 3 into Scopes 1 and 2 under GH1. And the aggregation model now weights assets by how long they were actually owned during the reporting year, according to GRESB’s own methodology notes.
Infrastructure participants get a parallel set of changes, including an optional NZIF Alignment Module that produces a net zero alignment report alongside the standard benchmark report. Recurring reporting gets lighter where nothing material has changed at an asset, with prefilled data and reusable validation findings from prior years.
Embodied carbon now needs a receipt
The embodied carbon change is the one worth slowing down for, and not because of the points.
Under the 2026 Real Estate Standard, evidence supporting an embodied carbon disclosure must now be uploaded by the participant under DMA2. T1.2 no longer accepts multiple net zero targets per entity. GH1 requires participants to go back and restate prior-year emissions data on the new scope basis.
That’s a different kind of ask. Reporting a policy is cheap. Producing a document that a validator will read, for a metric most of the industry only started measuring recently, isn’t. Embodied carbon sits in materials and construction rather than in a meter reading, so the data tends to live with contractors and suppliers instead of with the asset manager filing the assessment. The 2026 Standard moves a slice of the assessment from what a manager says it does toward what a manager can show, and once a benchmark starts asking for receipts it can’t easily stop.
Why reclassify tenant-space emissions?
Alignment with the GHG Protocol, which is the unglamorous answer and the correct one.
Landlord-controlled tenant space energy was being reported as Scope 3, which put GRESB participants in the position of reconciling their benchmark submission against their statutory and voluntary climate disclosures every year. Moving it into Scopes 1 and 2 removes that reconciliation. It also means restating history, since the change applies to prior years as well as the current one.
The numbers are smaller than they look
The calibration shows up in the percentages. The GH1 reclassification touches roughly 3.1% of operational assets. The ownership-period weighting materially affects around 2%.
GRESB’s stated aim is to “reduce the reporting burden for participants and streamline the reporting process over time, while enabling the reallocation of points to higher-priority issues.” Read against those percentages, that reallocation looks carefully sized: enough to move the incentive, small enough that no participant can credibly claim the ground shifted under them mid-cycle.
The governance question underneath the methodology
Standard-setting bodies fail in two directions. Move too slowly and the benchmark stops meaning anything to the investors who use it. Move too fast and participants can’t supply what you’re asking for, scores become noise, and the comparison the whole exercise depends on breaks.
Rose’s own background sits on the regime-building side of that problem. She is a founding member of GFANZ, the Glasgow Financial Alliance for Net Zero, and her government work has run through the UK’s Energy Efficiency Taskforce and the Net Zero Council. Those are all bodies that set expectations which institutions then have to meet, and all of them have run into the question the 2026 Standard is answering: how much evidence can you ask for before the ask outpaces the industry’s ability to supply it?
The scale involved explains the caution. The 2025 cycle drew 2,382 real estate assessments from 1,002 fund managers and 805 infrastructure assessments from 186 managers. Average standing investment scores reached 79, and energy data coverage was recorded above 75% globally. More than 80% of participating entities in both asset classes now report a net zero policy. Those are the numbers a methodology change has to survive contact with.
The same benchmark also shows why raising every bar at once isn’t available. Only 22% of assets in the 2025 cycle met ASHRAE Standard 100 for energy efficiency. On a metric like that, a scoring change would mostly sort participants by which building stock they happen to own rather than by anything they decided, which is the failure mode a benchmark administrator is paid to avoid.
Rose holds the seat as Independent Non-Executive Chair, a governance role rather than an operational one, and her stated read of what the organisation is for is narrow in a useful way. GRESB, she said on taking the role, “plays a critical role in equipping investors and managers with data and insights that drive responsible long-term growth.” The operative word is equipping. A benchmark administrator that began dictating outcomes would lose the thing that makes its numbers worth comparing in the first place.
The 2026 Standard suggests the answer GRESB settled on. Raise the evidence bar where the industry can already clear it, retire what nobody was learning anything from, and leave the harder measurement questions for a cycle when more than a handful of participants can answer them.
