ACC globe logo Affect Climate Change Inc.Education. Evidence. Practical action.
Sustainability CompanionCompanion Donate

AI & data-center charities

Stand with communities.
Help reduce AI’s harm.

AI’s environmental impacts need action. ACC stands with people organizing to stop harmful data-center projects. Find groups pushing for accountability, measuring impacts and working toward less damaging technology, and see how to support them.

Find a group to support Understand the issue
Illustration of community members examining plans for a proposed data center.
Community voices belong at the center of these decisions.Illustration; not a photograph of a listed organization.

Where ACC stands

Communities have the right to say no.

Pausing, redesigning or rejecting a harmful data-center project can be the right outcome. Making AI more sustainable also means reducing unnecessary use, demanding independent scrutiny and securing enforceable protections for water, air, energy costs and local decision-making.

AI is one part of the data-center sector. The groups below take different approaches and do not all oppose new development. Technical improvements must be judged alongside the harms a project may still cause.

Organizations you can support

Turn concern into action.

Start with the kind of work you want to support. These five giving and sponsorship links go directly to the recipient; each group’s role and limits are set out below.

Protect communities & demand accountability.

Community organizing and independent policy work help people challenge harmful expansion.

Measure impacts & reduce emissions.

These tools can inform changes to computing and electricity use. Their results depend on methods, assumptions and implementation.

Examine infrastructure solutions critically.

Compare proposed improvements, industry relationships and evidence of results before deciding whether to support this work.

Gifts support these organizations’ broader work unless the recipient confirms a restriction. Giving does not offset AI use, prove a project is sustainable or replace community consent. Listings and logos identify the organizations; they do not establish an ACC partnership or endorsement by the listed groups.

Follow the work. Check the claims.

A little context before you give.

Read how the tools work, what ACC has committed to, and which wider resources can help you investigate further.

Our 1% commitmentMultiple charities
Multiple charities · First direct contribution pending

We commit a total of 1% of eligible earned revenue across multiple charities working to make AI more sustainable. ACC donations and grants keep their designated purposes. Recipients and allocations may vary. The first direct contribution under this updated policy is pending; completed gifts will be distinguished from amounts set aside. No partnership or endorsement is implied.

Read the commitment and calculation →
How the CodeCarbon tools work

Inside CodeCarbon

How the tools work

CodeCarbon Computing on accessible hardware
Uses available hardware energy readings and estimates to calculate electricity use. It combines that result with electricity carbon-intensity data to estimate greenhouse-gas emissions.
Read the CodeCarbon method →
EcoLogits AI running on remote servers
Models supported AI requests using model and usage information, plus assumptions about infrastructure. It estimates indicators such as energy, emissions and water; it does not directly meter the provider's servers.
Read the EcoLogits method →

Their published methods make the assumptions and limits visible, helping people interpret the estimates and look for ways to reduce impacts.

Featured organization roles checked October 28, 2026. Giving destinations retain the October 21 review; recipient terms can change. Review each recipient’s current checkout settings before confirming a gift.

The wider resource directory

Research, standards & practical tools.

These groups work on research, policy, software standards and infrastructure. Their roles and support options differ. Listing a resource does not make it our 1% recipient.

The full directory covers 15 organizations and projects. Some work specifically on AI; others address the wider digital sector. Not all offer public charitable giving or an AI-specific fund.

Climate research & education

Climate Change AI

Connects machine learning with climate research, education and practical applications. Explore its resources on when machine learning can help, its limits and its emissions footprint.

ACC’s support remains planned while we confirm how to give. No direct checkout is listed here.

Explore its work ↗Giving information ↗
Browse the wider directory

Research, policy and digital sustainability

  • Ceres

    Publishes guidance on sustainable data centers, covering energy, water, grid strain, and community impacts beyond AI alone.

    Its work covers data centers broadly. Listing Ceres here does not select it as our 1% recipient.

  • Green Web Foundation

    Builds tools for fossil-free web hosting and digital carbon accounting across the internet, not AI specifically.

    Its wider digital mission is not an AI-specific cooling or water-efficiency program.

  • Data & Society

    Examines climate technology and justice, studying how tech infrastructure affects environments and communities.

    Research can inform community decisions; this is not a facility retrofit program.

  • Epoch AI

    Tracks data on AI compute growth and data center capacity, offering resource estimates and methodology.

    Power-capacity estimates are not metered electricity or water use, and research does not itself establish reductions.

  • AlgorithmWatch

    Produced the SustAIn project, publishing AI sustainability assessments and policy recommendations.

    The original SustAIn project ended in 2024; its published resources remain available for reference.

Standards, benchmarks and engineering

  • Green Software Foundation

    Develops the Software Carbon Intensity for AI and other green-software standards and methodologies.

    Standards guide measurement; applying them does not by itself demonstrate lower resource use.

  • MLCommons

    Runs MLPerf Power, benchmarking energy use of AI systems under standardized testing conditions.

    System benchmarks do not measure whole-facility impacts or establish cooling-water savings.

  • AI Energy Score

    A separate collaborative project rating inference energy use of AI models under controlled tasks and hardware.

    Inference GPU energy is not the same as total facility energy, training impacts or water use.

  • Open Compute Project

    Publishes open data center facility designs covering power, cooling, and water engineering approaches.

    Engineering resources do not establish verified savings from a particular project or donation.

Wider directory descriptions reviewed October 15, 2026; featured organization roles rechecked October 28, 2026. Listings are informational and do not establish an ACC partnership, contribution or chosen 1% recipient.

Return to the AI education discussion →

How ACC is changing its own AI use.

Alongside supporting community action, we’re testing tools and reducing unnecessary AI requests. See what we’ve tried, what providers claim and what we have yet to measure.

Our AI Use & Tools →