Google DeepMind Launches AI Accelerator for Climate in Asia-Pacific

Google DeepMind is launching its inaugural Accelerator program in Asia Pacific, dedicated to what the company calls "AI for the Planet." The announcement targets one of the starkest regional mismatches in the global climate effort: a region driving global economic growth while remaining highly exposed to climate change.

The program runs for three months and opens with an in-person bootcamp in Singapore. It accepts startups, research teams and nonprofits from across the Asia-Pacific region. Google says selected organizations will work on problems in nature, climate, agriculture, energy and related domains using frontier AI.

The stakes are concrete. Google's announcement points to a recent report showing that green technologies are gaining momentum but are not scaling fast enough to keep up with the region's rising environmental risks. That gap — between the pace of technological adoption and the pace of environmental deterioration — is the problem the program is designed to attack.

What participants get

Organizations accepted into the accelerator will receive three forms of support, according to Google's announcement:

• Expert mentorship from Google staff.
• Tailored support for their specific projects.
• Help integrating frontier AI and science AI models from Google AI experts into their products or research.

The third element is the program's distinguishing feature. Rather than offering general-purpose startup advice, Google is positioning the accelerator as a channel for direct technical collaboration — giving regional teams working access to Google's frontier and science-focused AI models as building blocks for environmental applications.

The stated target audience is deliberately broad. Startups, academic or independent research teams, and nonprofits all qualify. That breadth matters for environmental work in Asia Pacific, where some of the most consequential climate interventions — in agriculture, energy systems and natural ecosystems — are often led by public-interest research groups and civil society organizations rather than venture-backed companies.

Google frames the invitation plainly. "If you're working on climate solutions, we want to help you scale your work," the announcement reads. Interested organizations can register their interest now, ahead of the Singapore bootcamp that kicks off the three-month program.

Why Asia Pacific

The company's rationale for anchoring the program in the region rests on two facts stated in the announcement. First, Asia Pacific is described as "a global engine for economic growth." Second, the region is "highly vulnerable to climate change."

Those two facts together define the regional problem the accelerator addresses. Rapid economic growth in Asia Pacific drives both emissions-intensive development and the capital needed to decarbonize it. At the same time, the region's exposure to climate impacts — which the announcement identifies as rising — creates pressure for solutions that arrive faster than conventional green technology adoption is currently delivering.

The recent report cited by Google, showing green technologies falling behind the region's environmental risks, gives the program its urgency. The announcement does not name the report, but its central finding is unambiguous: current scaling trajectories are insufficient.

Frontier AI meets environmental science

The program's scope — "nature, climate, agriculture, energy and more" — signals that Google is aiming the accelerator at applied scientific problems, not just software products. The reference to "science AI models" alongside frontier AI suggests participants will be able to draw on models built for scientific domains, a category Google DeepMind has invested in heavily.

For research teams and nonprofits in the region, access to frontier AI infrastructure and Google experts is the program's core offering. Building advanced AI capability in-house is expensive; the accelerator's model transfers that capability to organizations whose missions are environmental rather than commercial AI development.

For startups, the program offers a faster route from climate-domain expertise to AI-powered products. The mentorship and integration support cover the gap between knowing an environmental problem and deploying frontier models against it.

What to watch

The inaugural run will test whether a three-month, mentorship-and-model-access format can meaningfully accelerate environmental AI work in a region as large and varied as Asia Pacific. Google has committed to an in-person launch in Singapore, a physical anchor that suggests the company intends the program to build regional community, not just distribute technical resources remotely.

Organizations working on climate, nature, agriculture or energy solutions in the region can register interest now; the bootcamp in Singapore opens the program. If the first cohort produces deployable AI applications that outpace the green-technology adoption curve the cited report describes, the accelerator's model — pairing frontier AI access with environmental mission teams — could become a template Google extends elsewhere.