Google DeepMind has selected 16 organizations to participate in the first cohort of the
Accelerator: AI for the Planet (APAC) program, with the goal of scaling AI-powered environmental solutions across the Asia-Pacific region. The cohort includes startups, nonprofits, and research teams working in biodiversity conservation, sustainable agriculture, and carbon-solution measurement.
The program began this week with a practical boot camp in Singapore, and participants will receive three months of support through specialized mentorship, technical assistance, and access to the Google AI suite, including customized advanced models. Google said participants will use models such as AnthroKrishi, ForestCast, AlphaEarth Foundations, SpeciesNet, and Perch to develop and scale their solutions.
Three Tracks for Environmental Applications
In the nature conservation and climate resilience field, the cohort includes 800 Trust, Listening Lab, and Wildlife.ai from New Zealand; Kumi Analytics, SIGMA, and City Syntax Lab from Singapore; as well as TelePIX from South Korea and Yayasan Ekosistem Lestari from Indonesia. Their applications range from bioacoustic analysis and biodiversity monitoring to using satellite data to track mangrove trees, building open-source cameras for wildlife conservation, and linking environmental degradation to disaster risks.
The sustainable agriculture track includes Edufarmers, Living Roots, SIGMA, Terrastack, and X-Centric. These organizations are working to provide farmers with near-real-time guidance on pests, diseases, and weather; design crop-specific biofertilizers; estimate productivity using satellites; provide field-level information; and use portable X-ray and AI-powered devices to analyze soil properties.
The third track focuses on climate and carbon solutions, with participation from Archeda in Japan; Climitra Carbon, Farmers for Forests, and Varaha Climate in India; and City Syntax Lab in Singapore. The projects include converting nature-based carbon credits into measurable assets, improving energy and carbon consumption in urban areas, verifying the removal of invasive species and converting them into biochar, and enabling smallholder farmers to measure regenerative agriculture practices and carbon credits.
What Matters in Practice?
The initiative is significant because it brings together AI models with satellite data, bioacoustics, and field devices, rather than limiting environmental uses to general analytical tools. The diversity of participants also shows that potential beneficiaries are not only technology companies; smallholder farmers, conservation projects, researchers, and organizations that measure carbon all fall within the program’s scope.
However, the announcement confirms support and model access for three months, and does not yet provide operational results or figures on emissions reductions or productivity improvements. The program’s actual value will therefore depend on the teams’ ability to move from prototypes to scalable applications, as well as on data quality and independent verification of the results of carbon and biodiversity solutions.