Tasmania’s government and Sustainable Timber Tasmania partnered with Indicium Dynamics and RoboticsCats to deploy AI-driven wildfire detection technology, enabling early detection and rapid response during the 2023–2024 fire season, helping to reduce the annual burned area to 20,000 hectares, less than half the prior decade’s average. Pano AI uses networks of rotating high-definition and infrared cameras to detect wildfires early, covering a 10-mile radius per station, cutting response times by over 20 minutes.87 Embrace the Forest (Brazil) under the wider Embrace the Forest initiative, umgrauemeio, in partnership with other private sector actors, government agencies, non-governmental organizations (NGOs) and Indigenous communities, operates 11 AI detection towers across 2.5 million hectares in Brazil’s Pantanal. This collaboration combines prevention and rapid response, reducing burned areas by 40% compared to 2020 during the severe 2024 fire season. FireSat, a partnership led by Earth Fire Alliance with Google. org, the Gordon and Betty Moore Foundation and Muon, is a satellite constellation designed for rapid wildfire detection. Scanning every 15–20 minutes, it can detect fires 400 times smaller than current systems and track them through smoke and darkness in near real time. In California alone, FireSat could prevent 190,000–350,000 acres from burning each year and avoid 4–8 million tonnes of CO2 emissions. CASE STUDY 3 Technology, AI and data for wildfire management Design principles for scalable wildfire data, technology and governance ecosystems These principles describe what empirical evidence suggests is likely needed to turn detection-to- response technologies and risk data into trusted, interoperable systems that can be governed, adopted and scaled. Not every principle applies to every case. –Reference standards: Use common data formats, quality checks, licences (open where possible) and a common alert protocol. –Last-mile operational integration: Deliver role-specific, decision-ready insights into the tools and workflows where they’re used – fast enough to change outcomes (e.g. alerts on wind shifts and fire behaviour). –Privacy and data rights: Build data systems that protect privacy and respect community and Indigenous ownership of information, while supporting responsible data use for risk reduction and underwriting. –Pilot-to-scale: Support flexible rules and funding for proven tech (drones and AI systems) to move to scale, and change management to facilitate integration (i.e. data, processes, training, labour agreements, etc.). Putting the pillar into practice TABLE 5 Sample outcome metrics Partners Contracts Policy enablers Economic and financial (e.g. change in EAL), human, social and health (e.g. detection latency, suppression response minutes), cross-sector and regional (e.g. mitigation feature inventory and status)Counties/HOAs, insurers, utilities, vendors, insurers, universities and incident agenciesData-sharing agreements, third-party assurance, procurement for integrated detect-to-respond systemsRules for drones/autonomous flight, privacy templates, international alert protocols and Wildfire Science & Technology Commons participation From Wildfire Risk to Resilience: The Investment Case for Action 22

From Wildfire Risk to Resilience The Investment Case for Action 2026