Published

1 year ago

June 2, 2025

Alice Rush

Bezos Earth Fund, founded by billionaire Jeff Bezos, founder and former CEO of Amazon, launched a major initiative called the AI for Climate and Nature Grand Challenge in April 2024. The program pledges up to $100 million to support teams using artificial intelligence (AI) to solve environmental problems. Recently, it revealed its first grantees or recipients of the fund.

The funding initiative focuses on real-world solutions. It aims to reduce carbon emissions and protect wildlife using smart technology. The goal is to link AI experts with environmental groups. This helps them use AI to solve tough climate and nature problems.

Many of these organizations have strong ideas but lack the tech expertise or funding to apply AI. This is where the Bezos Earth Fund comes in, offering grants and encouraging teamwork across fields.

The challenge focuses on four main areas:

Sustainable proteins. Finding AI tools that speed up the discovery and production of plant-based or alternative proteins.
Biodiversity conservation. Using AI to track endangered species, protect ecosystems, and stop threats like illegal logging.
Power grid optimization: Developing smarter, cleaner ways to store and distribute renewable energy.
Wildcard innovations. Supporting creative AI ideas that don’t fit into a standard category but have strong environmental potential.

Let’s get to know who these grantees are and what they do.

Grantees Tackling Carbon Removal and Climate Solutions

In May 2025, the Bezos Earth Fund announced its first 24 grantees, each receiving $50,000 to build out their ideas. Some of the most impactful focus on climate mitigation and carbon removal, including:

Carbon Sim: CO₂ Removal Accelerator (Yale University)

This project uses AI-powered simulations to test and improve strategies for carbon dioxide removal (CDR). It aims to help scientists quickly evaluate which methods—like soil enhancement or ocean capture—are the most effective at storing carbon.

EV Charging Optimization (Cornell University)

Cornell’s team is creating a tool that uses real-time AI to manage charging for electric vehicles (EVs). It adjusts when and how cars are charged so they act as energy storage for the power grid. This can support the shift to renewable energy and help reduce emissions.

Livestock GPT

Another Cornell project, Livestock GPT is a generative AI tool that helps dairy farmers cut methane emissions. It includes a chatbot that gives feed and farm advice—especially for use in emerging economies—helping reduce climate-warming gases from livestock.

These climate-focused grantees aim to tackle emissions directly while making climate solutions more scalable and accessible.

Why This Matters for Climate and Nature

AI has the potential to supercharge global environmental efforts—but only if it’s applied wisely and equitably. The AI for Climate and Nature Grand Challenge is helping turn that potential into reality by:

Giving environmental groups access to cutting-edge AI tools
Funding early-stage ideas with clear pathways to impact
Encouraging partnerships between tech and nature experts
Supporting scalable, verifiable, and science-backed solutions
Helping meet global climate targets faster and more affordably

The need for innovation in climate and nature solutions has never been greater. According to the Intergovernmental Panel on Climate Change (IPCC), global greenhouse gas emissions must be cut by nearly 50% by 2030 to keep warming below 1.5°C and avoid the worst impacts of climate change.

At the same time, the World Economic Forum estimates that over $44 trillion of economic value—more than half of global GDP—is moderately or highly dependent on nature and its services, underscoring the stakes for biodiversity loss.

The Power of AI in Climate Action

AI is increasingly recognized as a game-changer for environmental action. A 2023 report by Boston Consulting Group found that AI could help reduce global greenhouse gas emissions by up to 10%—the equivalent of 2.6 to 5.3 gigatons of CO₂e—by 2030, if deployed at scale across sectors like energy, transport, and agriculture.

Key AI applications to accelerate climate progress

Yet, a 2022 survey by Microsoft and Goldsmiths University revealed that only about 43% of environmental organizations felt “AI-ready”. They cited barriers such as lack of funding, technical expertise, and access to data.

Bridging the Gap: The Role of the Grand Challenge

The Bezos Earth Fund’s AI for Climate and Nature Grand Challenge directly addresses these barriers by providing critical funding and technical support to early-stage projects. By awarding $50,000 seed grants to 24 diverse teams in its first round, the initiative is lowering the entry threshold for nonprofits, universities, and startups to experiment with AI-driven solutions.

This approach is vital, as early-stage funding for climate tech remains scarce—just 6% of global venture capital in 2023 went to climate-related startups, according to PwC. And in 2024, VC deals for climate tech innovations further drop from 2023, per Pitchbook data.

climate tech investment 2024 — Source: Pitchbook

The Grand Challenge also fosters collaboration between AI experts and environmental practitioners, a proven recipe for success. For example, projects like Carbon Sim (Yale) and Livestock GPT (Cornell) are bringing together machine learning specialists, ecologists, and farmers to co-design tools that are both scientifically robust and practical for real-world use. Such partnerships help ensure that solutions are not only technologically advanced but also grounded in local knowledge and needs.

AI in Action: Use Cases Beyond Carbon

Other grantees use AI to help the environment. They reduce food waste, create better plant-based proteins, and protect forests. Here are some of them and their innovations that attracted Bezos Fund’s investment:

University of Leeds – Food Waste to Protein. This project uses AI to turn food waste into sustainable protein. The software finds the best microbes and fermentation methods.
Wageningen University – OLiMPuS Platform. This open-source AI platform helps scientists find new plant and fermented proteins that feel and taste like milk and meat.
BGCI-US – Forest Monitoring and Illegal Logging Detection. Using drones and AI, this project tracks over 500 endangered timber species and detects illegal logging in real time.
AI-powered Forest Monitoring in the Amazon. Another grantee is working in the Amazon rainforest, combining satellite data with AI to detect changes in forest cover.
AI for Coral Reef Health (University of Miami). This project uses underwater cameras and AI models to assess the health of coral reefs. It can detect bleaching and pollution damage early.

AI isn’t just about crunching data. It’s also a strong tool for early detection, quick decision-making, and scaling nature-positive solutions.

Scaling Up: What Happens Next?

The $50,000 planning grants are just Phase I. Later in 2025, up to 15 teams will move to Phase II, receiving $2 million each over two years to scale and implement their solutions. This will allow them to move beyond prototypes and test their tools in real-world settings.

The Bezos Earth Fund says it’s also building partnerships with AI labs, tech companies, and universities to support the technical side of the challenge. At the same time, it wants to train environmental groups on how to use and trust AI, ensuring that solutions are not only powerful but practical.

The projects supported by the Bezos Earth Fund are still in early stages, but they point toward a future where smart software can support a healthy planet. Whether it’s managing forests, cleaning up farms, or inventing new kinds of food, AI is now part of the climate and conservation toolbox.

The post Bezos Earth Fund’s AI for Climate and Nature Reveals First Grantees appeared first on Carbon Credits.

Carbon Footprint

How to improve Scope 3 data accuracy for CSRD

Published

7 hours ago

July 17, 2026

Alice Rush

For most businesses, the emissions that matter most sit outside their own walls. Scope 3 emissions, everything generated across your value chain, from the suppliers who make your inputs to the customers who use your products, typically make up the majority of a company’s total carbon footprint. Under the Corporate Sustainability Reporting Directive (CSRD), those value-chain emissions now have to be measured and disclosed with a rigour that spend-based estimates alone struggle to satisfy. This guide sets out how to improve Scope 3 data accuracy for CSRD: the calculation methods open to you, how to move from estimates to verified supplier data, and how to govern that data so it holds up to audit.

Carbon Footprint

How community stewardship makes carbon credits durable

Published

14 hours ago

July 17, 2026

Alice Rush

A carbon credit is a commitment that extends well into the future. The tonne of CO₂ compensated for today from a nature-based carbon project must remain out of the atmosphere for good, which means the forest behind the credit has to remain standing long after the transaction is complete. For any buyer, this raises a defining question: What ensures that the forest endures?

Carbon Footprint

Why Conventional Carbon Offsets Are Losing Boardroom Credibility

Published

2 days ago

July 15, 2026

Alice Rush

What replaced the cheap REDD credit on the boardroom slide deck, and why procurement is leading the rewrite.

Three years ago, a corporate slide showing a portfolio of cheap REDD+ credits could carry a board meeting. The number was big, the price was low, and the press release wrote itself. Today, that same slide gets sent back with questions. The questions are uncomfortable, the answers are unclear, and your general counsel is suddenly in the room.

Conventional carbon offsets are not dead. The voluntary carbon market retired 202 million tonnes in 2025, and the Morgan Stanley Institute for Sustainable Investing survey published in January 2026 confirmed that interest from corporate buyers remains substantial. What changed is the credibility threshold. The integrity floor has risen, the disclosure scrutiny has tightened, and the buyer profile has shifted. This article tracks what changed, what sophisticated buyers now ask before signing, and what serious corporates are putting on the board slide instead.

What boards used to buy, and why it stopped working

The 2020 to 2022 model was simple: buy a large tranche of avoidance credits at low single-digit prices, retire them against the company footprint, announce the carbon-neutral claim, and move on. Most of those credits came from REDD+ projects, renewable energy installations in countries where the renewable energy was already economic, or methane projects with thin documentation.

Several things broke that model. Academic research published in 2023, including a widely cited Science paper, found that the majority of REDD+ credits issued under the most common methodologies did not represent additional reductions when tested against rigorous counterfactuals. The Voluntary Carbon Markets Integrity Initiative published its Claims Code of Practice, which sets requirements for what companies can credibly claim from credit use. The European Union finalised its Green Claims Directive, restricting how companies can describe products as climate-neutral. France’s Décret 2022-539 already restricts carbon neutrality advertising. California’s AB 1305 imposes disclosure requirements on any company making net-zero or carbon-neutral claims while doing business in the state.

The collective effect: the cheap credit no longer buys the announcement, and the announcement now carries litigation risk.

The integrity reset: ICVCM, VCMI, and what changed

The Integrity Council for the Voluntary Carbon Market published the Core Carbon Principles in 2023 and began assessing methodologies against them in 2024. The first methodologies received the CCP label later that year. The point of the label is to give corporate buyers a defensible quality screen they can cite in disclosure.

The Voluntary Carbon Markets Integrity Initiative complements this on the demand side. Its Claims Code of Practice defines what a buyer can say (Silver, Gold, or Platinum claims, with associated requirements) based on the quality of credits used and the underlying decarbonisation strategy. Together, CCP and VCMI build a quality stack: CCP on the supply, VCMI on the claim, with the science-based target sitting underneath both.

The reset is not a ban on offsets. It is a ratchet. Credits that meet the new bar continue to clear; credits that do not, do not. The Morgan Stanley survey found that 61% of current buyers like the CCP label concept but that supply of labelled credits remains limited. That supply constraint is now visible in pricing.

What sophisticated buyers ask before they sign

The questions on the procurement scorecard have changed. A 2022 buyer might have asked about price, vintage, and project type. A 2026 buyer asks five different questions before any of those.

What does the counterfactual look like, and who validated it.
What is the permanence regime, and what is the buffer pool exposure.
What is the leakage risk, and how is it mitigated.
What rating has the project received from the independent ratings agencies (Sylvera, BeZero, Calyx Global), and what was the rationale.
What is the documentation discipline that survives an audit four years from now when the procurement team that signed the contract has moved on.

If the vendor cannot answer those five questions on a first call, the conversation ends. Conversely, if the vendor can answer them with documented specificity, the conversation often expands beyond a single transaction toward a multi-year engagement.

Where this leaves your near-term commitments

You probably have near-term commitments that pre-date the integrity reset. Public targets to be carbon neutral by 2025 or 2030. Product-level claims that ran in last year’s marketing. Disclosed reduction trajectories that assumed continued access to cheap credits.

You have three workable paths. The first is to re-baseline your strategy, replacing the most exposed credits with higher-quality alternatives and adjusting the public language to match what you can defend. The second is to shift the underlying spend from offsetting outside your value chain to investing inside your value chain, where reductions count against Scope 3 directly and the audit trail is cleaner. The third is to keep the strategy and absorb the risk, which is increasingly the most expensive option once you price in litigation, restatement, and reputational exposure.

Most serious buyers are choosing the second path. It moves the carbon spend from a compliance cost to a procurement and resilience investment, and it removes the central failure point of the legacy model: the disconnect between where the emissions occurred and where the reductions sat. Nature-based supply chain investments, structured under the GHG Protocol Land Sector and Removals Standard and aligned to the SBTi FLAG Guidance, are the asset class that fits this brief. They generate inventory-grade reductions, they produce audit-grade documentation, and they survive the new claim restrictions because the carbon math sits inside the value chain that the disclosure already covers.

If you are reassessing a carbon strategy under the new integrity bar, or rebuilding a board narrative that has to survive a more skeptical audience, the carbon and sustainability experts at Carbon Credit Capital can help. The Dual-Value Model gives you a defensible alternative to legacy offset purchases, with the documentation and operational integration that survives the procurement scorecard and the audit. Schedule a consultation.