For more than two decades, organizations have been encouraged to set decarbonization (or net-zero) targets and use carbon accounting mechanisms to track emissions and monitor their progress toward those goals. However, net-zero targets are difficult to reconcile with growth ambitions, and the decarbonization benefits of carbon accounting are becoming harder to demonstrate, leading to a crisis in carbon accounting.

Net-zero goals

As the link between greenhouse gas (GHG) emissions and climate change has become clear, organizations and society have sought to measure and control emissions, often seeking to decarbonize energy systems and eventually balance out all the emissions they cause — a so-called net-zero goal.

These goals should be validated by a third-party audit conducted in accordance with ISO 14068. The Science Based Targets initiative (SBTi), established in 2015, is the leading organization for validating net-zero goals.

Around 14,000 organizations have validated their targets with SBTi or committed to doing so, a number that has risen over the past few years but remains a small fraction of companies worldwide. Net-zero targets represent around one-third of the targets set by participating companies. The SBTi's dashboard shows a steady increase in net-zero targets set, from around 1,600 in March 2025 to 2,800 in September 2026.

Compared with other sectors, data center operators are well represented on the dashboard. Many large colocation operators, such as Digital Realty and Equinix, have SBTi-validated net-zero targets. Other large players, including the major hyperscalers — AWS, Google, Microsoft and Meta — have set near-term targets with SBTi or self-validated net-zero goals.

Carbon accounting

Leading regulatory mechanisms for reducing emissions include pricing GHG emissions and implementing carbon taxes. To comply with these systems and measure progress toward net-zero goals, organizations need to inventory their carbon emissions, a process known as carbon accounting.

This requires carbon accounting standards; the most widely used is the Greenhouse Gas Protocol (GHG Protocol), which classifies emissions as Scope 1 (direct), Scope 2 (indirect) and Scope 3 (supply chain).

Since the late 1990s, organizations have set GHG emissions reduction goals (sometimes validated by SBTi) and measured their progress toward them by managing emissions inventories using the GHG Protocol. These organizations have reduced Scope 1 and 2 emissions by using carbon-free energy (CFE), paying for new CFE sources through power purchase agreements (PPAs), or purchasing energy attribute certificates (EACs) from existing CFE generation.

Scope 3 emissions are a significant problem. They are often cited as representing around 90% of a company's total emissions, but Scope 3 reporting is vastly more complex, duplicative and uncertain than Scope 1 and 2, resulting in sparser, lower-quality reporting.

Goals for Scope 1 and 2, backed by carbon accounting, have had considerable success, driving a large increase in CFE use through PPAs. Large technology firms account for 30% of all corporate PPAs, supporting 77 GW of renewable power capacity.

Limitations of carbon accounting

Despite the successes of carbon accounting and net-zero targets, this approach faces two types of issues: implementation problems and limitations on the results it can achieve.

Implementation challenges

• Inconsistent validation and enforcement. Corporate net-zero targets often remain unvalidated. The Uptime Institute Global Data Center Survey 2026 found that 50% of operators have net-zero targets (see Uptime Institute Global Data Center Survey 2026). An earlier survey, conducted in 2023, found that around 40% of targets were validated through SBTi; others may have been validated by third parties. 
  Carbon accounting is mandatory for large data center operators under the EU's Corporate Sustainability Reporting Directive and California's Senate Bill 253, but optional elsewhere. Mandatory reporting requirements are emerging in Brazil, Hong Kong, Japan, New Zealand, Singapore, Switzerland and the UK. Scope 3 reporting requirements vary between jurisdictions.
• Low adoption and poor data. Operators report data unreliably or inconsistently, and there is no standardized approach to allocating Scope 2 emissions from colocation providers to their tenants. Uptime Institute's 2026 annual survey found that 60% of organizations do not collect Scope 1 emissions data, 69% do not collect Scope 2 emissions data and 79% do not collect Scope 3 emissions data.
• Complexity. The three Scopes create confusion and discourage reporting. Despite the GHG Protocol, standards and guidance offer alternative accounting approaches, making it difficult to understand and compare goals across the data center industry.
• Scope 3 estimates are particularly ambiguous. SBTi requires a Scope 3 target if Scope 3 accounts for more than 40% of an organization's emissions (i.e., almost all organizations), but these targets are based on poor-quality data. Emissions estimates for categories such as other energy-related emissions, embedded emissions in products and materials, and waste disposal carry uncertainties of 50% or more. SBTi allows Scope 3 targets to be less ambitious than those for Scope 1 and 2. Uptime Intelligence research shows that only a minority (21%) of data center operators report Scope 3 emissions.
• Double counting. Ambiguity over how CFE generation is credited means, for example, that colocation operators and their clients may offset the same emissions multiple times.

Limited results

• Limits of local generation. Under location-based accounting, the GHG Protocol balances Scope 2 emissions against locally generated CFE, but net zero cannot be achieved in this way in most grid regions. CFE from wind and solar requires battery storage to cover hourly emissions and, even with batteries, can economically achieve only 80-90% local CFE consumption. Achieving zero emissions from the grid requires reliable, dispatchable CFE sources such as nuclear and geothermal. 
  Market-based accounting sidesteps this limitation by allowing data center operators to match local consumption with EACs for energy used on other grids. Generation will not be synchronized with consumption in either space or time and may lead to surplus generation and curtailment during periods of high output.
• Financial pressure. Organizations are under pressure to proritize profitability. This can overshadow demanding goals to decarbonize or achieve net zero by 2030. A rapid increase in energy consumption, for example, requires fossil-fuel sources such as natural gas and increases demand for carbon credits, which are becoming scarcer and more expensive. Purchasing EACs or offsets to cover Scope 2 emissions can also be an expensive proposition.
• Carbon accounting neglects infrastructure. Company-specific actions typically do not address gaps in transmission infrastructure, which can prevent new CFE sources from connecting to demand. The inability to export stranded CFE can impede progress toward global or national decarbonization.

The constraints of increased precision

Both SBTi and the GHG Protocol are being rewritten (see Scope 2 Guidance update: impact on climate disclosure), but the revisions may address implementation issues while overlooking the system's limitations. New standards and draft guidance from SBTi and the GHG Protocol have sparked fierce debate over purity and purpose, with the two organizations appearing divided over the future direction of corporate climate action. Proposed fixes generally tighten requirements by expanding mandatory regimes and introducing penalties for non-compliance.

In an attempt to make offsets more effective, a draft update to the GHG Protocol proposes imposing hourly matching on EACs used in market-based accounting. This solution distances carbon accounting from the real world, where the available data is inadequate to support hourly matching.

While the GHG Protocol attempts to tighten the accounting rules governing the underlying tools, SBTi has taken a somewhat more practical course in the first update to its Net Zero Standard since its launch in 2015. SBTi v2 allows a "best-efforts" framework, gives organizations of different sizes and in different locations greater flexibility, and replaces fixed long-term targets with optional goals on a 5-year rolling cycle. It also embeds and formalizes carbon credits rather than rejecting them.

SBTi v2 will be open for submissions from February 1, 2027. SBTi v1.3.1 will remain open until the end of January 2028, at which point all new submissions must use SBTi v2. Validations last for 5 years, so all SBTi v1 targets will expire by the end of February 2033.

SBTi declined to explicitly align v2 with the draft update to the GHG Protocol's Scope 2 guidance. Hourly matched targets are not required, but organizations using more than 10 GWh per year will be asked to report the percentage of their consumption that is matched hourly.

Both bodies have downplayed their differences in joint statements and public events. At the London launch of SBTi v2, GHG Protocol chief executive Tim Mohin called for less polarization, and SBTi chief executive David Kennedy said: "There's nothing in the SBTi Corporate Net Zero Standard that is at all inconsistent with [GHGP's] direction of travel toward market instruments."

The EU considered adopting a 15-minute, time-matched guarantee of origin for renewable energy consumption (similar to the draft GHG Protocol) in its rating scheme and key performance indicator (KPI) for data centers (see Draft EED delegated regulation sidesteps critical issues). Following lobbying by the sector, the proposal was omitted from the Delegated Regulation, which was published in September 2026 and is expected to enter into force after a 2-month scrutiny period by the European Parliament and the Council.

The International Organization for Standardization (ISO) is currently consulting on a Net Zero standard aligned with SBTi v2, which may attempt to accommodate both approaches. The outcome cannot be foreseen; it is unlikely to resolve differences among the standards and certifications, and may err on the side of overly restrictive accounting requirements.

Underlying problems

These efforts do not address the limitations discussed above. Net-zero targets and carbon accounting drive a limited range of corporate actions.

Efforts to offset Scope 1 and 2 emissions have driven investment in new renewable energy projects and reduced the price of low-carbon electricity. However, in markets with high penetration of intermittent CFE, curtailment levels are increasing, making additional CFE less effective.

In carbon accounting, organizations balance their books in isolation through short-term, low-cost purchases of clean energy generation and/or EACs. Electricity grids now need new storage capacity and transmission infrastructure to better integrate this generated electricity. These investments do not produce easily defined reductions in tons of GHG emissions and cannot easily be packaged into instruments such as EACs.

At this point, more precise accounting addresses a largely solved problem — making low-carbon generation affordable — and distracts from the system-level reforms required to enable more effective use of that generation.

Alternative ways forward

Scope 2 Guidance: If the GHG Protocol retains hourly matching as a requirement of the market-based method, the data center sector, led by the hyperscalers, will likely propose an alternative standard that maintains the market-based accounting approach permitted under the current Scope 2 guidance.

Uptime Intelligence believes that a suitable level of accuracy can be achieved more simply by using a monthly tally of energy consumption and average emission factors (see Scope 2 Guidance update: impact on climate disclosure).

Scope 3 Guidance: Efforts to improve Scope 3 accounting will continue but as with the Scope 2 update process, are likely to become bogged down in intractable arguments. Uptime Intelligence believes that reporting organizations can ultimately do little to reduce Scope 3 emissions and that responsible organizations should address them through Scope 1 and Scope 2 reductions.

Alternate methodologies: Ending the drive for accounting precision might allow attention to shift to the more fundamental need for a framework that directs investment toward system-level projects that reduce Scope 1 emissions (ultimately, the only emissions that matter, since every emission is someone's Scope 1 emission). This may require metrics other than tons of GHG emissions.

One possibility is impact accounting (also known as consequential accounting or emissionality), promoted by the Corporate Energy Buyers Association (CEBA), which measures real-world reductions of GHG emissions from the electricity grid (see Impact accounting for corporate energy buyers). CEBA advocates for investment in grid infrastructure, storage projects, nuclear restarts and tariff innovation alongside traditional PPAs. However, impact accounting focuses on Scope 2 emissions, leaving the issue of Scope 3 untouched.

Another approach is from the Task Force for Corporate Action Transparency (see Mitigation action accounting and reporting guidance). This aims to incentivize actions that change the amount of carbon in the atmosphere, but it measures all actions in tons of carbon, so its effect on shared infrastructure may likewise be limited.