Closing the Gap: How NCEAS Is Using AI to Unlock the Full Potential of Biodiversity Monitoring

graphic with different automated digital asset collecting devices

Have you ever uploaded a photo to iNaturalist or used an app like Merlin ID to identify a bird call? When you use these tools, you’re helping create digital assets: data products that can be used to monitor the incredible diversity of life on Earth.

Accurate, timely biodiversity data are essential for protecting species and ecosystems. Yet much of the information needed for effective conservation still arrives too late, at too coarse a scale, or with critical gaps. At the National Center for Ecological Analysis and Synthesis (NCEAS), researchers are using artificial intelligence (AI) not to replace field science, but to help close these gaps, ensuring biodiversity monitoring can keep pace with the accelerating rate of environmental change.

As part of our AI for the Planet initiative at NCEAS, the Digital Assets for Biodiversity project is a collaborative effort advancing automated biodiversity monitoring through improved digital tools, shared frameworks, and AI-powered workflows. The project includes recent NCEAS-led research showing that while automated monitoring technologies hold enormous promise, their implementation still lags far behind what is needed for effective global conservation.

a bird becoming digitized

The Promise (and the Reality) of Automated Biodiversity Monitoring

In a 2025 paper led by NCEAS researcher Rachel King, the team conducted the first global assessment of automated biodiversity monitoring systems, cataloging more than 250 existing “digital assets” used to track living organisms worldwide. The findings were striking: most digital assets rely on satellite data, primarily monitor plants at coarse taxonomic resolution, and are heavily concentrated in North America and Europe. Even when data are collected frequently, delays in processing often mean they become available months or years later, limiting their usefulness for real-time decision-making.

“These technologies already exist, but we’re not fully realizing their potential,” says King. “AI can help us move from raw data to usable information faster but only if the workflows, infrastructure, and access are in place.”

The Digital Assets for Biodiversity project at NCEAS is designed to do exactly that: identify bottlenecks, fill critical gaps, and accelerate the responsible use of AI in biodiversity monitoring.

Figure 1. A framework laying out the process of digital asset creation to define and address bottlenecks in their production. The framework divides creation of digital assets into four distinct steps or Investment Areas (gray boxes), and each area has potential Methods (yellow boxes) for addressing a bottleneck in the creation of digital assets at that step and some potential Benefits (purple boxes) of investing in that area are listed to highlight the usefulness for different monitoring needs. The benefits shown are not exhaustive but represent relevant gains from investing in improving a particular area. A comprehensive list of benefits for each investment area can be found in table S3. — Figure 1 from King et al. 2025, caption adapted: A framework laying out the process of digital asset creation to define and address bottlenecks in their production. The benefits shown are not exhaustive but represent relevant gains from investing in improving a particular area.

What Are Digital Assets and Why AI Matters

When someone uploads photos or recordings to an app like iNaturalist or Merlin ID, they are contributing to a much larger system of digital assets. In the Digital Assets for Biodiversity project, digital assets include data products generated by automated or semi-automated technologies such as camera traps, acoustic sensors, radars, and hydrophones as well as community science platforms that use similar AI-enabled approaches, see figure below.

Together, these tools can collect biodiversity data at scales impossible for traditional field surveys. But that scale comes with a challenge: they generate enormous volumes of images, audio recordings, and sensor data that would be impractical to process manually.

This is where artificial intelligence becomes essential. Machine learning algorithms can identify species from photos or sound recordings, extract ecological indicators from satellite imagery, and integrate data across platforms. Tools like iNaturalist use AI to suggest species identifications, enabling both scientists and the public to contribute meaningfully to biodiversity monitoring at scale.

“Before AI, someone might have spent months manually labeling images or recordings,” explains King. “Now, AI dramatically reduces that workload and shortens the time between data collection and conservation action.”

At NCEAS, the focus is on using AI as a tool to increase speed, accuracy, and transparency — while ensuring that digital assets remain reproducible, well-documented, and fit for real-world decision-making.

figure 2 A diversity of automatic recorders to monitor ecological communities non-invasively and remotely. (1) Vocalising birds being monitored by microphones deployed on trees. (2) Stridulating and drumming fishes being recorded by hydrophones attached to moorings. (3) Walking elephants producing ground vibrations perceived by geophones. (4) Fish shoal being detected by a sonar. (5) Oceanic glider navigating an Environmental Sampling Processor (ESP) to sample eDNA. (6) Bear being detected by camera traps fixed on trees. (7) Hyperspectral camera mounted on a drone and monitoring tree composition in a forest. (8) LiDAR sensor mounted on an unmanned aerial vehicle monitoring multiple forest canopies. (9) Imaging flow cytometer attached to a mooring and recording planktonic communities. (10) Racoons being detected by thermal and IR cameras at night. (11) Stationary radar and a satellite radar, respectively, monitoring bird and large mammal populations. Recorder's ability to detect the presence of living organisms, count their numbers, classify them at the species level and measure their traits (e.g. behavioural, functional and morphological traits) is evaluated from 1 to 3 levels as follows: 1 bar corresponds to ‘in corner-case situations only’, 2 bars corresponds to ‘in specific conditions and on specific organisms (for detecting, counting and classifying) or for a limited number of features (for measuring)’, and 3 bars corresponds to ‘in most cases and for most organisms (for detecting, counting and classifying) and for several features (for measuring)’. — Figure 2 from Besson et al 2022, caption adapted: A diversity of automatic recorders to monitor ecological communities non-invasively and remotely. (1) Microphones to record bird calls. (2) Hydrophones to record stridulating and drumming fishes. (3) Geophones to record vibrations from walking elephants. (4) Sonar to detect fish shoals. (5) Oceanic glider to sample water for traces of organisms. (6) Camera traps to take photos of bears. (7) Hyperspectral camera (on drone) to monitor tree composition in a forest. (8) LiDAR sensor (on an aerial vehicle) to monitor forest canopies. (9) Imaging flow cytometer to record planktonic communities. (10) Thermal and IR cameras to detect raccoons (night). (11) Stationary radar and satellite radar, respectively to monitor bird and large mammal populations. Recorder's ability to detect the presence of living organisms, count their numbers, classify them at the species level and measure their traits.

Using AI to Address Real Gaps in Conservation Data

The 2025 NCEAS-led study identified several major challenges limiting the impact of automated biodiversity monitoring:

Taxonomic gaps, especially for insects, microorganisms, and many marine species
Geographic biases, with far fewer monitoring assets in biodiversity-rich regions
Processing bottlenecks, where data exist but are not transformed into usable products
Delays in data availability, reducing relevance for management decisions

The Digital Assets for Biodiversity project responds directly to these findings. By improving AI-driven data processing workflows, the project helps convert raw sensor data into accessible monitoring products more efficiently, one of the highest-impact leverage points identified in the research.

AI in Practice: Building the Future of Digital Assets

AI is already reshaping how NCEAS researchers collaborate and synthesize information, and its value was on full display during the NCEAS-hosted workshop, “Innovation in Digital Assets for Biodiversity.” The workshop brought together experts across disciplines, technologies, and regions to examine gaps in automated biodiversity monitoring and identify practical paths forward.

During the workshop, King used AI tools to rapidly extract and organize ideas from hundreds of handwritten sticky notes, saving hours of manual work and allowing participants to focus on solutions rather than logistics. Beyond facilitation, AI is also helping researchers streamline coding, data cleaning, and documentation by supporting reproducible, open-source workflows that can be applied across species and regions.

“These tools don’t replace expertise,” King notes. “They free us up to focus on synthesis, interpretation, and collaboration: the things humans do best.”

Workshop discussions connected digital assets to global conservation priorities, including Essential Biodiversity Variables, ecosystem monitoring frameworks, and international targets such as 30x30. Case studies highlighted how AI-enabled tools from camera traps to wildlife re-identification systems can support more timely, inclusive, and actionable biodiversity assessments.

group shot from DA workshop — Group photo of the Innovation in Digital Assets for Biodiversity Workshop attendees and facilitators.

Responsible AI, Transparent Science

Across all NCEAS projects, responsible AI use is non-negotiable. Researchers document where and how AI is used, clearly describe workflows, and make code and data publicly available whenever possible. Ultimately, AI is treated as a tool, not an authority.

“AI can speed things up, but researchers are still responsible for checking results and explaining how conclusions are reached,” says King. “Transparency in AI use is what makes these tools trustworthy.”

Looking Ahead: AI for Faster, Fairer Conservation Decisions

Looking forward, NCEAS researchers see AI playing a growing role in making biodiversity data easier to find, interpret, and apply. Emerging approaches such as plain-language data querying and edge computing (where AI processes data directly on sensors or drones) could reduce delays, lower barriers to access, and enable near–real-time monitoring in remote regions.

As the Digital Assets for Biodiversity project makes clear, automated biodiversity monitoring already holds substantial promise. By combining AI with synthesis, collaboration, and open science, NCEAS is helping turn that promise into practical tools for conserving life on Earth: faster, more equitably, and at the scale the challenge demands.

an earth with data and wildlife image overlay