Google DeepMind has selected 16 startups, nonprofits and research teams for its inaugural Accelerator: AI for the Planet (APAC) program, a three-month initiative designed to help organizations develop and scale AI-enabled environmental tools.
The cohort begins with a bootcamp in Singapore and will receive access to Google’s AI stack, including specialized frontier models, alongside tailored technical support and mentorship. The group covers organizations across Australia, India, Indonesia, Japan, New Zealand, Singapore, South Korea and Thailand.
The focus is environmental measurement
The cohort is notable less for a single application than for its concentration on turning fragmented physical-world information into usable decisions. Many participants rely on satellite imagery, bioacoustics, sensors, drones and agronomic data—the inputs needed to monitor systems that are difficult and expensive to observe directly.
In conservation, New Zealand’s 800 Trust and Listening Lab are using bioacoustics for biodiversity monitoring, while Wildlife.ai is building open-source AI-powered camera systems. Singapore’s Kumi Analytics combines remote sensing with deep learning to establish conservation baselines. South Korea’s TelePIX is applying satellite data to mangrove monitoring, and Indonesia’s Yayasan Ekosistem Lestari is developing a predictive platform linking environmental degradation to disaster risk.
For operators, the underlying opportunity is continuous monitoring rather than intermittent surveys. That could improve the speed at which conservation groups, land managers and public agencies identify threats—provided models can perform reliably across local geographies and field conditions.
Agriculture and carbon are key commercialization paths
Five cohort members are focused on agriculture. Indonesia’s Edufarmers delivers pest, disease and weather guidance to smallholder farmers through messaging apps. Thailand’s Living Roots uses data and AI to formulate crop-specific biological fertilizers. Singapore-based SIGMA is developing satellite AI models for crop-yield estimates; India’s Terrastack provides plot-level land intelligence; and Australia’s X-Centric combines portable X-ray hardware with AI for geochemical analysis.
The common business challenge is distribution. Advice and measurement are valuable only if they fit existing farmer workflows, work at low cost and produce outcomes that users can verify. Messaging apps, portable hardware and plot-level analysis point to different approaches for narrowing that last-mile gap.
Carbon and climate-accounting projects form another significant share of the cohort. Japan’s Archeda is working to make nature-based carbon credits measurable at scale using satellite data. India’s Varaha Climate and Farmers for Forests are targeting carbon and biodiversity measurement for smallholder farming and agroforestry, while Climitra Carbon uses Geo-AI to verify invasive-species removal and biochar conversion. Singapore’s City Syntax Lab is building an agentic AI platform to optimize urban-district energy and carbon.
These projects address a central constraint in climate markets: credible measurement, reporting and verification. AI can reduce the cost of processing remote-sensing and field data, but its commercial value will depend on whether buyers, auditors and regulators accept the resulting evidence.
What to watch over the next three months
Google DeepMind says participants will use frontier models and receive expert mentorship to navigate technical complexity. The important test will be whether that support produces deployable systems rather than stronger demonstrations.
Executives should watch for three signals: integrations with field-data and customer workflows; evidence that models generalize beyond pilot sites; and independent validation of environmental or financial outcomes. For founders in climate tech, the cohort also reinforces where large AI platforms see demand: domain-specific systems built around proprietary or hard-won environmental data, not generic AI features.
The accelerator does not by itself validate the organizations’ products or establish their impact. But it gives 16 APAC teams a route to technical resources and visibility at a time when environmental intelligence is becoming increasingly important infrastructure for agriculture, conservation, cities and carbon markets.
