Search Climate Insights Directory

The report analyzes 65 state-level tariffs to identify mechanisms that protect ratepayers from the costs of surging demand from large energy users. It highlights five key safeguards: minimum contract terms, minimum monthly billing demands, collateral requirements, exit fees, and capacity reassignment. The author argues that while these tariffs provide a foundation for fair cost allocation, they should be paired with broader strategies like grid flexibility and optimized planning to ensure reliability and affordability.

The policy brief examines how virtual power plants (VPPs) can mitigate grid interconnection delays for large loads, specifically data centers, in the United States. It highlights that VPPs can scale to meet over 20% of US peak demand by 2030 and are significantly cheaper than new gas plants. The document proposes three commercial models—pass-through funding, capacity transfer, and reliability reinforcement—to align incentives between large loads, utilities, and VPP aggregators. It also recommends reforms in interconnection policy, integrated planning, and customer protection to enable this transition.

PJM is facing a capacity crisis driven by surging demand from data centers and a slow interconnection process that takes over eight years to bring new generation online. This delay has caused capacity market prices to skyrocket, with the total capacity bill rising from $2.2 billion in December 2022 to $16.1 billion in the latest auction. The author identifies long study timelines and high transmission network upgrade costs as primary barriers, noting that PJM has connected fewer new resources relative to peak demand than ERCOT or CAISO. To fix this, the document recommends implementing third-party automation software, reforming Energy Resource Interconnection Service (ERIS) to be a faster pathway, integrating advanced transmission technologies (ATTs), and prioritizing proactive regional transmission planning.

The report identifies a conservative USD 4.4 trillion pipeline of creditworthy sustainable bonds from 12 climate-leader countries through 2035. It argues that global megatrends—such as electrification, urbanisation, and the rise of AI—will further accelerate this pipeline beyond current projections. The document also outlines policy levers, including taxonomies and monetary policy, that governments can use to channel capital into mitigation and resilience projects.

The report critiques the 'firm fixation' approach where utilities meet surging data center demand by adding gas plants or refurbishing coal units. It identifies six defining features of data centers—agency, clustering, consumption profiles, flexibility, backup requirements, and modularity—that make them unsuitable for simple one-to-one matching with firm power plants. Instead, the author advocates for a 'system of systems' approach combining clean energy portfolios, battery storage, and demand-side flexibility. Key recommendations for policymakers include using the interconnection process as leverage to ensure cost causation and exploring demand-side resources, such as sponsoring household energy upgrades, to free up grid capacity.

The briefing reveals that twenty major private equity firms are linked to nearly 60 GW of operating gas-fired power capacity globally, emitting 82 million metric tons of CO2e annually. It highlights a growing trend where private equity firms invest in both AI data centers and the fossil fuel power plants required to run them, deepening carbon lock-in. While the US holds the largest share of this capacity, private equity ownership is highly concentrated in countries like Togo and Panama.

The report reveals that twenty major private equity firms are linked to nearly 60 gigawatts of operating gas-fired power capacity globally, emitting an estimated 82 million metric tons of CO2e annually. This fleet spans 27 countries, with significant concentrations in the United States, Togo, Panama, Jamaica, and the Philippines. The document highlights a growing financial convergence where private equity firms are investing in both AI data centers and the fossil fuel power plants required to run them, potentially deepening carbon lock-in. Specific examples include BlackRock's pursuit of AES Corp and Blackstone's acquisition of TXNM.

The report argues that Brazil's digital infrastructure development is characterized by a 'narrative of progress' that obscures ecological damage and marginalizes affected communities. Through media analysis, community surveys, and official information requests, the study finds a systematic lack of environmental impact assessments and a significant gap between the government's 'green' rhetoric and the reality of resource consumption. In Rio Grande do Sul, the Scala AI City project is situated in a flood-prone area, while in Ceará, the TikTok-partnered project by Casa dos Ventos faces opposition from the Anacé Indigenous community over land and water rights. The study concludes that the 'burden of the cloud' falls disproportionately on vulnerable populations through a combination of regulatory vacuum and strategic disinformation.

Private equity firms have invested nearly $200 billion in data center-related industries since January 2022, driving a strategy of acquiring vertically integrated retail energy utilities and fossil fuel power plants to secure power for AI infrastructure. BlackRock's Global Infrastructure Partners (GIP) has acquired ALLETE Inc. and is pursuing a $38 billion takeover of AES, while Blackstone has invested in the Potomac Energy Center and is seeking to acquire TXNM Energy. The report argues that this expansion risks creating stranded high-emitting assets if AI demand is overestimated, while already causing increased electricity costs for ratepayers—rising up to 267% in some data center hubs—and significant environmental degradation. Organized community resistance has already delayed or blocked $64 billion in data center projects across 24 U.S. states.

The World Association for Christian Communication (WACC) warns that AI infrastructure is following an extractivist economic model, citing environmental damage and lack of local benefit. It specifically highlights a planned US$25 billion OpenAI data centre in Argentina as an example of this trend. WACC advocates for national oversight, the protection of Indigenous rights, and a moratorium on new data centres until water and energy risks are mitigated.

In 2025 Q3, US utility resource plans showed increasing projected electricity demand and emissions, with many utilities delaying fossil plant retirements to ensure resource adequacy. Load growth is primarily driven by large loads like data centers, while the phase-out of federal renewable tax credits and MISO regulatory changes have led to reduced wind and solar plans and increased reliance on gas capacity. Current IRP projections indicate a significant gap between planned emissions reductions and the net-zero targets required by 2035 for climate alignment.

The document outlines the climate and community risks associated with the proliferation of AI data centers, noting their high energy and water demands and their tendency to rely on fossil fuels. It provides practical advice for journalists to report on local data center capacity, utility rate increases, and pollution, while citing specific examples of community struggles in the US and AI's varied roles in climate adaptation and forecasting.

Thailand's economic strategy to become a regional hub for electric vehicles (EVs) and data centres requires a rapid transition to clean energy to attract foreign investment and meet international low-carbon standards. The energy sector, which accounted for 69% of total emissions in 2019, is central to Thailand's goal of reducing greenhouse gas emissions by 30-40% by 2030 and achieving carbon neutrality by 2050. Projections indicate that EVs could account for 20% of total electricity demand by 2037, while data centre demand is expected to grow at an 8% CAGR from 2025 to 2027, potentially adding 10 TWh by 2037. To support this, the Thai government has introduced a direct power purchase agreement (DPPA) pilot project specifically for data centres of at least 50 MW, allowing them to bypass the single-buyer system and purchase renewable energy directly.

The report proposes a cost-optimal energy pathway for Thailand that exceeds the Revised Power Development Plan (RPDP) targets by adding 32 GW of solar and 6 GW/15 GWh of battery storage by 2037. This shift would result in cumulative cost savings of $1.8 billion and avoid 147 million tonnes of CO2 emissions. The analysis suggests that Thailand can meet its demand without adding 2 GW of planned natural gas capacity, as solar and batteries can cover growth from data centres and EV charging. While wind power is deemed financially unviable unless capital expenditures drop by 10% annually, the report suggests importing hydropower from Lao PDR as a cost-effective alternative to wind, despite risks related to import dependency and climate-driven droughts.

The report proposes a cost-optimal energy pathway for Thailand through 2037, suggesting that increasing solar capacity by 89% and battery storage by 60% relative to the draft revised power development plan (RPDP) 2024 can save $1.8 billion in power generation costs. This pathway aims to meet rising electricity demand from data centres (projected 10 TWh increase) and electric vehicles (projected 3.3 GW peak load increase) while reducing natural gas imports by 1,815 billion cubic feet. The analysis finds that wind energy is not financially viable without a 10% annual CAPEX reduction, suggesting imported hydropower from Lao PDR as a more cost-effective alternative.

The document examines the 'twin transition' of digital and green transformations, highlighting how digital technologies can both enable climate action and contribute significantly to global greenhouse gas emissions. It identifies critical gaps in infrastructure and financing for developing countries and proposes five priority decarbonization levers: green infrastructure, energy-efficient operations, sustainable supply chains, circular economy models, and climate-compatible infrastructure. The authors argue for coordinated international cooperation, targeted financial incentives, and the integration of environmental provisions into digital trade agreements to ensure an inclusive transition to net zero.

The Union of Concerned Scientists (UCS) reports that utilities in seven PJM states passed over $4.3 billion in data center connection costs to all customers in 2024. This occurs because large data centers (100–1,000 MW) require high-voltage transmission infrastructure, which falls under federal FERC jurisdiction and is typically allocated across all customers, unlike low-voltage distribution connections. UCS finds that in 95% of cases, these costs are not paid by the data center owners. The brief recommends that FERC and state PUCs require utilities to track and assign these specific costs to the customers causing them.

The briefing challenges the notion that coal plants are necessary for grid reliability amidst a data center boom in the US. It asserts that coal plants are prone to unplanned outages, lack the ramp speed required for voltage-sensitive loads, and create grid congestion. The author suggests that over 95 percent of future demand can be met using scalable clean solutions such as battery storage and demand flexibility.

As of Q2 2025, US utility resource plans show a trend of increasing electricity demand projections and a corresponding increase in planned gas power plant capacity, leading to a widening gap between projected emissions and net-zero targets.

The 2025 Corporate Climate Responsibility Monitor finds that most major global companies fail to align their business models with the Paris Agreement's 1.5°C goal. While some companies like Apple, H&M Group, and Stellantis show 'moderate' integrity, none of the 20 companies in the sector deep dives achieved 'reasonable' or 'high' integrity ratings. The report argues that GHG emission reduction targets alone are insufficient and must be complemented by transition-specific alignment targets to ensure credible decarbonisation.