SAN FRANCISCO, July 21 — Two billion. That is the number of species occurrence records now housed in the Global Biodiversity Information Facility, an intergovernmental initiative launched in 2001. The figure that matters for anyone tracking the business of environmental data: those records come from 1,700 data publishers across 62 countries.
By the numbers, citizen science biodiversity databases have become the backbone of modern species monitoring. The bottom line: these platforms now generate the raw feedstock for tens of thousands of scientific papers every year.
Investors in conservation technology and environmental consulting will note that the data pipeline has grown far faster than anyone predicted at the turn of the century. iNaturalist, co-founded by Ken-ichi Ueda and Scott Loarie in 2008, became a joint venture of the California Academy of Sciences and National Geographic Society in 2014. As of 2023, the platform had logged 100 million observations from 3 million users. eBird, launched in 2002 under the Cornell Lab of Ornithology, had accumulated over 1 billion bird sightings from 600,000 users by the same year.
GBIF now holds over 2 billion occurrence records. Those are the headline numbers.
Here is the catch: the same open submission model that allows these platforms to scale so quickly also makes them vulnerable to fake data. The challenge has intensified since the rise of generative AI.
The AI infiltration problem
The first known AI-generated images in iNaturalist were reported in late 2022. By mid-2023, the number had grown to hundreds. Researchers and platform operators are now grappling with a question that has real consequences for data integrity: how do you keep a system open to millions of contributors while filtering out synthetic submissions?
For context, iNaturalist’s 100 million observations and eBird’s 1 billion sightings did not accumulate by accident. These platforms rely on volunteers uploading photos, audio recordings, and location data.
The verification process traditionally depends on community review — other users confirm or challenge identifications. Generative AI undermines that trust at scale because synthetic images can look convincing to human reviewers.
Established infrastructure, emerging risk
GBIF was established in 2001 as an intergovernmental initiative. It now represents one of the largest aggregators of biodiversity data on the planet, drawing from institutional collections as well as citizen science platforms. The scale is extraordinary: 2 billion occurrence records, 1,700 publishers, 62 countries.
Tens of thousands of scientific papers draw on this data annually. eBird has been running since 2002, predating the smartphone era. Its 600,000 users have submitted over 1 billion bird sightings.
Cornell Lab of Ornithology runs the operation, and the dataset has become a standard reference for ornithologists, conservation planners, and land managers. iNaturalist, with its 100 million observations and 3 million users, has become the default platform for general biodiversity observation. The vulnerability in these systems is structural. Open submission is what made them succeed.
Closing that openness would cut off the very pipeline that produces the data. The question facing database operators is whether automated detection tools can keep pace with generative AI.
What to watch next
The trajectory is clear: generative AI imagery in biodiversity databases appeared in late 2022 and grew to hundreds of instances within months. The data publishers themselves — 1,700 institutions across 62 countries contributing to GBIF alone — will need to decide whether to invest in detection infrastructure or accept a certain level of noise in the record. For researchers publishing tens of thousands of papers annually from this data, the margin for error matters.
Sources
{“@context”: “https://schema.org”, “@type”: “NewsArticle”, “headline”: “The Rise of Citizen Science Databases: From iNaturalist to GBIF”, “description”: “SAN FRANCISCO, July 21 — Two billion. That is the number of species occurrence records now housed in the Global Biodiversity Information Facility, an intergovernmental…”, “datePublished”: “2026-07-21T09:30:00+00:00”, “author”: {“@type”: “Person”, “name”: “Ava Anderson”}, “publisher”: {“@type”: “Organization”, “name”: “InfoPulse Today”, “url”: “https://newsfeed.infopulsetoday.com/”}, “citation”: [{“@type”: “CreativeWork”, “url”: “https://en.wikipedia.org/wiki/Citizen_science”}]}
