Google DeepMind Selects 16 Environmental AI Projects for First APAC Accelerator
Google DeepMind has chosen 16 organizations from 8 APAC countries for its first “AI for the Planet” accelerator cohort, running through December 2026.
Photo by Alesia Kozik on Pexels
Google DeepMind announced on September 7, 2026, the first cohort of its “AI for the Planet” accelerator for the Asia-Pacific region, selecting 16 startups, nonprofits and research teams from a competitive applicant pool. The organizations span Australia, India, Indonesia, Japan, New Zealand, Singapore, South Korea and Thailand, and will use Google AI models to work on climate, agriculture and nature-protection projects over a three-month program.
The accelerator opened with a week-long in-person bootcamp in Singapore, running September 7 to 11, 2026, according to the program page on deepmind.google. Participants then move into virtual mentorship from September through December 2026, guided by a dedicated relationship manager, before the program concludes with an in-person Demo Day in December where teams present to mentors, investors and industry partners.
Applications for the program closed on July 26, 2026. The accelerator’s own program page states Google DeepMind planned to select “10-15 organizations” for the cohort; the final group announced numbers 16, one more than the top of that stated range.
What the selected organizations do
Google grouped the 16 organizations into three focus areas. Six work on protecting nature and building climate resilience: 800 Trust, Listening Lab and Wildlife.ai, all based in New Zealand, along with Singapore’s Kumi Analytics, South Korea’s TelePIX and Indonesia’s Yayasan Ekosistem Lestari. Their projects include bioacoustic monitoring of biodiversity, AI-powered wildlife cameras, remote-sensing baselines for conservation, satellite-based mangrove monitoring, and a predictive platform linking environmental degradation to disaster risk.
Five organizations focus on sustainable agriculture: Indonesia’s Edufarmers, which sends pest, disease and weather guidance to smallholder farmers through messaging apps; Thailand’s Living Roots, which designs biological fertilizers for specific crops; Singapore’s SIGMA, which builds satellite-based models to estimate crop yields; India’s Terrastack, which combines satellite and agronomic data for plot-level land intelligence; and Australia’s X-Centric, which replaces traditional soil labs with portable, AI-enabled X-ray hardware.
The remaining five target climate and carbon solutions: Japan’s Archeda, which uses satellite data to convert nature-based carbon credits into measurable assets; Singapore’s City Syntax Lab, which is building an agentic AI platform to optimize energy and carbon use across urban districts; and three India-based ventures – Climitra Carbon, which uses Geo-AI to verify invasive-species removal converted into biochar; Farmers for Forests, which uses AI-powered drones to measure carbon and biodiversity outcomes from smallholder agroforestry; and Varaha Climate, which uses remote sensing and AI to verify regenerative farming and carbon removal for smallholder farmers.
| Country | Organizations selected |
|---|---|
| India | 4 (Terrastack, Climitra Carbon, Farmers for Forests, Varaha Climate) |
| New Zealand | 3 (800 Trust, Listening Lab, Wildlife.ai) |
| Singapore | 3 (Kumi Analytics, SIGMA, City Syntax Lab) |
| Indonesia | 2 (Yayasan Ekosistem Lestari, Edufarmers) |
| Australia | 1 (X-Centric) |
| Japan | 1 (Archeda) |
| South Korea | 1 (TelePIX) |
| Thailand | 1 (Living Roots) |
The AI models behind the program
Google said participants will get access to its AI stack, including specialized frontier models, tailored support and mentorship from company experts. The blog post naming the cohort specifically calls out five models that organizations will draw on.
| Model | What it does |
|---|---|
| AlphaEarth Foundations | An AI model that integrates petabytes of satellite, radar and climate data into a unified “embedding” for each 10×10 meter patch of Earth’s land and coastal waters, supporting mapping of deforestation, agriculture and urban change |
| SpeciesNet | A model used to identify wildlife species, including from camera-trap imagery, to support biodiversity monitoring |
| Perch | An open bioacoustics model, available on Kaggle, that analyzes audio recordings such as bird calls and underwater hydrophone data to help conservationists track endangered species |
| ForestCast | A model described by Google Research as moving from counting historical forest losses to predicting future forest risk |
| AnthroKrishi | Google DeepMind’s agricultural AI initiative, built around its Agricultural Landscape Understanding (ALU) and Agricultural Monitoring & Event Detection (AMED) models, which originated in India and now also cover six African countries |
Program benefits also include equity-free support, training, technical project partnership with Google experts, and eligibility for cloud credits through the Google for Startups Cloud program or Google Cloud for Research, according to the accelerator’s program page. Graduating teams join what Google describes as its Accelerator Alumni Network, a global community it says spans more than 2,000 startups and nonprofits.
About Google DeepMind
Google DeepMind is Google’s AI research lab, formed from the 2023 merger of DeepMind and Google Brain. The lab has built a series of environmental AI tools ahead of this accelerator, including AlphaEarth Foundations, introduced in 2025 as a planetary mapping model now used by more than 50 partner organizations such as the UN Food and Agriculture Organization and Harvard Forest. Its Perch bioacoustics model, first released and then updated in August 2025, helps conservationists process audio data on endangered species including Hawaiian honeycreepers and coral reef ecosystems. The AnthroKrishi agriculture initiative began with India-focused models before expanding data coverage to Kenya, Uganda, Ghana, Rwanda, Nigeria and Zambia, and has drawn $2.5 million in Google.org funding for a related project with the FAO. The APAC accelerator is Google DeepMind’s first program of this kind in the region and puts these existing models directly into the hands of 16 organizations working on environmental problems across eight countries.
Sources: Google (The Keyword); Google DeepMind; TNGlobal; Carbon Herald; Google (The Keyword); Google DeepMind; Google India Blog
Featured image: photo by Alesia Kozik on Pexels (free Pexels license).
Become a Sponsor
Our website is the heart of the mission of WINSS – it’s where we share updates, publish research, highlight community impact, and connect with supporters around the world. To keep this essential platform running, updated, and accessible, we rely on the generosity of you, who believe in our work.
We offer the option to sponsor monthly, or just once choosing the amount of your choice. If you run a company, please contact us via info@winssolutions.org.
I specialize in sustainability education, curriculum co-creation, and early-stage project strategy. At WINSS, I craft articles on sustainability, transformative AI, and related topics. When I’m not writing, you’ll find me chasing the perfect sushi roll, exploring cities around the globe, or unwinding with my dog Puffy — the world’s most loyal sidekick.
