GBIF functions through dedicated teams, called nodes, that are situated in and act on behalf of countries and organizations that have chosen to become GBIF Participants. Becoming a Participant and establishing a node is the primary way that governments and international organizations contribute to GBIF. As of August 2026, GBIF has 70 Participant countries. Recognizing that comprehensive global data mobilization requires worldwide participation, it remains a priority to grow the GBIF network with the addition of new Participants. All Participants sign the GBIF Memorandum of Understanding (GBIF 2010), declaring three core intentions: (1) to establish a team to serve as a Participant node; (2) to commit to the principle of free and open access to biodiversity data; and (3) to actively participate in the governance and growth of GBIF.

Participants govern GBIF through the GBIF Governing Board, with voting rights held by countries that are Voting Participants. To maintain Voting Participant status, countries must agree to contribute financially to GBIF. National contributions follow a progressive funding structure and are framed as nonbinding, which has been instrumental in facilitating continuing commitments. National financial contributions are the main funding source for GBIF. The financial contributions are aligned to national GDP and renegotiated every five years along with the Strategic Framework.

Growth in funding from nations is primarily due to new Participants joining GBIF, rather than from increased contributions from current Participants. Following the recommendations in 2020 from the Committee on Data International Science Council (CODATA) to substantially increase core funding, GBIF has been able to leverage the core Participant contributions to double its operational budget with supplementary funding from private and public sector grants and contributions. The development and management of these new funding relationships has required new staffing and organizational structure which have additional costs.

Countries that are not yet Voting Participants can engage with GBIF through regional activities (see the “Capacity Development” section) and may join GBIF on a provisional basis as an Associate Country Participant, a status which is limited to five years and does not confer voting rights.

Participant nodes guide and implement the activities of GBIF in their respective countries or organizations and across GBIF as a whole. National nodes are typically hosted by an existing biodiversity-related institute in the Participant country, with flexibility to operate in a way that best serves the national and regional biodiversity data communities. Participant countries vary in their initial capacity to contribute, so the capacity of more experienced nodes can be leveraged to develop newer nodes. All nodes have core functions in common: they coordinate national and regional initiatives related to biodiversity data; they support mobilization of biodiversity data; they encourage the use of GBIF-mediated data for research and policy; and they act as a knowledge hub for biodiversity data (GBIF Secretariat 2025). Nodes provide technical expertise to leverage the GBIF informatics infrastructure, and many also support additional informatics services. In addition, many nodes contribute to national data sharing policies and protocols, and support the preparation of national biodiversity-related policy reports and assessments. Finally, each node contributes to GBIF governance through the Nodes Committee (GBIF governance is presented in further detail below). Across all functions, nodes aim to embody four key attributes: neutrality, leadership and initiative, a focus on service, and adaptability (GBIF Secretariat 2025).

Organizations that become GBIF Participants agree to the same core intentions as Country Participants and sign the GBIF Memorandum of Understanding, but their role in the GBIF community is somewhat different. Rather than focusing on a single nation, organizational nodes generally focus on either a specific thematic area or regional or subregional activity, often providing national nodes with support for thematic and regional activities. Though they do not have voting rights in the GBIF Governing Board, organizational Participants are encouraged to participate in GBIF governance discussions.

The biodiversity data made accessible by GBIF are published by many institutions around the world, referred to as GBIF data publishers. Data publishers are institutions involved in biodiversity data collection and management, including research institutions, museums, public agencies, academic journals, citizen science platforms, and, increasingly, private sector businesses that deal with biodiversity data in some capacity. As novel types of data and data uses emerge, GBIF continues to diversify its data publishing pathways to meet the needs of a growing data publisher community (Robertson et al. 2022, Sica et al. 2026). As of August 2026 there are over 2700 data publishers that have contributed at least one dataset to GBIF, with nearly 700 additional registered publishers that have not yet contributed a dataset; the number of registered publishers currently grows at a rate of more than two new publishers every three days.

The relationship between data publishers and GBIF is formalized through nodes’ endorsement of publishers, which serves to validate that data publishers are institutions or organizations with data holdings that are relevant to GBIF; that they have, or can be provided with, access to stable data-hosting services; and that they can provide a point of contact to handle feedback on the published data. The requirement that all data publishers provide a responsive point of contact is essential for publisher accountability and attribution and facilitates community-led curation and quality control of GBIF-mediated data. Furthermore, data publishers must consent to the data publisher agreement, including assignment of a Creative Commons license allowing for open reuse of the published data. In 2014, following a community-wide consultation, the GBIF Governing Board established a general policy to ensure that all datasets are associated with digital licenses by Creative Commons (CC0, CC BY, CC BY-NC). As of August 2016, all open-access species occurrence datasets on GBIF.org carry standardized, machine-readable licences. GBIF promotes attribution (CC BY) as the standard practice for citing GBIF-mediated data, reflecting an established norm to cite original work including data.

Though the majority of GBIF data publishers are data-holding institutions, there are two mechanisms that extend data publishing to individual data collectors who are not affiliated with an endorsed data publisher. First, researchers can publish a data paper in a peer-reviewed journal that is, or provides access to, an endorsed data publisher. Second, individuals can publish data to GBIF by sharing records through a GBIF-aligned citizen science platform.

GBIF provides technical support, primarily carried out by the Participant nodes, to lower the technical threshold for becoming a data publisher. The technical process of data publishing, including the development of publishing pathways to support emerging data types and new communities of data publishers, is described in the “How does GBIF work?” section.

The GBIF community includes anyone who uses GBIF-mediated data. As of July 2026, over a quarter of a million unique email addresses have been used to access data through GBIF, and GBIF-mediated data are downloaded at an average rate of 10 billion records per day. In terms of research usage, in 2026 GBIF recognized the milestone of enabling over 15,000 peer-reviewed publications based on GBIF-mediated data (GBIF Secretariat 2026c), and the geographic reach of GBIF-enabled research has grown to include nearly all the world’s countries (Steinke et al. 2025). GBIF supports the science-policy interface, both through scientific research that feeds into information products such as the IPBES assessments (Noesgaard et al. 2024) and through direct use of GBIF-mediated data in information products and indicators. Examples include the IUCN Red List assessments, the Species Information Index, and the Bioclimatic Ecosystem Resilience Index, which support national biodiversity reporting towards the Kunming-Montreal Global Biodiversity Framework, as well as decision-making through such mechanisms as National Biodiversity Strategies and Action Plans. Communication among GBIF nodes and the GBIF community, including helpdesk support and the GBIF community forum, indicate that GBIF-mediated data further support additional uses that are less easily tracked, including use in preliminary analyses, education, and local environmental planning and decision-making, among others.

Whichever way GBIF-mediated data are used, data users are considered—and, by consenting to the GBIF Data User Agreement, agree to become—a part of the broader GBIF community. As stated in the GBIF citation guidelines, “[d]ata accessed through the GBIF network is free for all—but not free of obligations” (GBIF 2017). Data users are asked to follow the data citation practices presented in the guidelines, including the use of a Digital Object Identifier (DOI). Users are also invited to notify GBIF of any concerns regarding data quality or other issues. In turn, GBIF provides support and capacity development for data users. In addition to support from GBIF nodes to data users in their regions or areas of focus, GBIF maintains online trainings, webinars, tools, and guidance to support a growing range of ways to query and access data. The ways that users can access GBIF-mediated data are described in the ”How does GBIF work?” section.

Much of GBIF's impact owes to the many individuals who make voluntary contributions to GBIF. Structured volunteer schemes drive GBIF’s work in certain areas, including translation of written materials, review and feedback of applicant proposals for GBIF-led programs, and mentoring and training across the GBIF network. Volunteers also play a key role in championing open data within biodiversity-related professional communities. GBIF’s Biodiversity Open Data Ambassadors program recognizes advocates for free and open biodiversity data and equips them with resources to increase the effectiveness of their advocacy. GBIF provides volunteers with capacity development opportunities as well as recognition and other professional benefits, and volunteers receive accreditation and recognition with digital badges awarded annually for active contributions through Open Badges. Beyond these structured roles, many individuals voluntarily contribute to the ongoing development of GBIF; much of the GBIF infrastructure, from software stack to written documentation to outreach activities, is open to community contribution, with documentation available on the GBIF website as an entry point for individuals who wish to get involved in these areas.

Collective decisions on all matters relating to GBIF are made by the Governing Board, which consists of the Head of Delegation from each Participant and meets once per year. Only Voting Participants have voting rights in the Governing Board, but representatives of Associate Country Participants, Other Associate Participants, and recognized observers are encouraged to participate in Governing Board discussions. The Governing Board steers the collective decision-making and strategic planning for GBIF, including the development of a Strategic Framework on a five-year cycle and an annual Work Programme. As GBIF reaches its twenty-fifth year, it is guided by the 2023-27 Strategic Framework, emphasizing growth in the GBIF community and expansion to new data types, informatics services, and data uses (GBIF 2022).

To advise and carry out specific responsibilities, the Governing Board has established an Executive Committee, which acts on behalf of the Governing Board when it is not in session, and three standing committees: the Science Committee, the Budget Committee, and the Participant Node Managers Committee. The Science Committee advises on implementation of the Work Programme. The Budget Committee provides oversight and risk assessment for the administration of GBIF funds. The Participant Node Managers Committee (often shortened to simply the Nodes Committee) consists of the node managers appointed by all Participants; its subcommittee, the Nodes Steering Group, consists of two node managers from each of GBIF’s geographic regions and three chair positions elected by the Governing Board and advises GBIF on behalf of the Nodes Committee. To ensure responsiveness to the broad communities served by GBIF, the committee at times establishes time-limited ad hoc Task Groups to bring together global experts on specific issues.

The GBIF Secretariat is charged with developing, executing, and reporting on the GBIF Work Programme. Housed in Denmark at the University of Copenhagen, the GBIF Secretariat employs 35 people as of 2026, in addition to contracted developers and regionally-based support teams. The primary tasks of the GBIF Secretariat include: operating GBIF informatics, including the GBIF portal; coordinating GBIF network activities related to data mobilization and publishing; supporting and monitoring the uses of GBIF-mediated data and identifying gaps; and operating GBIF’s communication and collaboration services. The Secretariat also facilitates the distributed execution of the Work Programme by diverse members of the GBIF community. Helpdesk services and informatics development are increasingly distributed across the GBIF network, with Participant nodes taking on significant leadership with the support of the Secretariat, and GBIF increasingly appoints regional contractors to guide and implement capacity development efforts in regions where there is not yet sufficient capacity through national nodes.

GBIF-mediated data derive from a wide range of data sources, which continues to expand as the technological capacity to accommodate complex data grows and GBIF develops ways to handle data associated with new communities of data publishers. As a result, the history of GBIF can be traced through phases marked by the emergence, integration, and growth of communities of data publishers, and the associated development of the GBIF infrastructure. The following paragraphs give a brief overview of some major data sources relevant to GBIF in the chronological order of their emergence, with the aim of offering a historical perspective on how GBIF has grown over time in response to the biodiversity data community (Fig. 5).

Natural history collections. In its earliest days, GBIF focused primarily on mobilizing specimens and other data from natural history museums and collections. To date, GBIF has mediated nearly a half billion digitized natural history records. This work remains a priority to this day (Fig. 5), as digitized records are still estimated to represent only a fraction of the world’s natural history specimens (Nelson and Ellis 2018, Hedrick et al. 2020). To increase the accessibility of natural history collections as the work of digitization continues, GBIF now maintains the Global Registry of Scientific Collections (GRSciColl), a hub of collections information originally developed by the Consortium of the Barcode of Life and formerly hosted by the Smithsonian Institution (Schindel et al. 2016, Grosjean et al. 2021). In GRSciColl over 270 million specimens have been linked to their collection and institution.

Incidental observation data. Non-vouchered data on scientific observations were another early data source for GBIF. This includes data derived from transcribed field books, scientific monitoring projects, or catch-and-release studies, among others. Participatory collection of data by amateur observers, broadly known also as participatory science (also often called citizen science), emerged in the 2010s as a major contributor of observation data to GBIF. Participatory initiatives continue to contribute a large amount of data, with several estimates indicating that around half of all data mediated by GBIF may be tied to participatory science, a substantial proportion of which derive from bird observation networks (Chandler et al. 2017, Heberling et al. 2021, Mandeville et al. 2023). The majority of these data have been published by participatory science platforms that are endorsed data publishers, enabling GBIF to ensure that there are sufficient quality control processes in place. In parallel with the growth of participatory science as a direct source of GBIF-mediated data, it has played a role in supporting the publishing of data from other sources, such as by supporting the transcription and digitization of data from museums and other natural history collections (Hedrick et al. 2020).

Molecular data. DNA-derived data emerged as a serious focus for GBIF around the mid-2010s and remain an area of growth today. Sequencing technologies, like DNA barcoding of specimens and metabarcoding of environmental DNA, enable researchers to detect, delimit, and quantify species and other taxa (Steinke et al. 2025). GBIF has partnered with DNA infrastructures, organizations, and experts in molecular research to integrate DNA-derived data into GBIF, increasing the data available in GBIF and highlighting the diversity of as-yet unclassified species, known as dark taxa. The first GBIF guide for publishing DNA-derived data was released in 2021 and subsequently updated, a Metabarcoding Data Toolkit was piloted in 2024 with continued development ongoing, and a new initiative on integration of eDNA data, SPLICE, was launched in 2026 (Abarenkov et al. 2023, GBIF Secretariat 2024b).

Survey data. Biological data from one-off surveys, repeating sampling and monitoring, and sensor-based surveys (e.g., camera and acoustic traps, biologging) represent the most active development area for GBIF. With the support of improved data standards and the first version of a guide for publishing survey and monitoring data to GBIF released in 2025, these data can now be shared in a way that better represents the survey design. This enables reporting and inference of organismal absence and abundance, facilitates improved data discovery, and increases the likelihood of data re-use in downstream statistical interpretation, metanalyses, and more (Ingenloff et al. 2025). GBIF is working within and beyond the existing publishing community to mobilize survey data from public and private institutions globally. As content layers of survey data continue to grow, it can be expected that the user base of GBIF will shift towards more time series research (e.g., trend analyses) to meet data demands in monitoring, management, and policy applications.

In parallel with these broad chronological phases, GBIF-mediated data have evolved along with broader trends in biodiversity research and informatics. The following paragraphs detail some key trends that have had an impact on GBIF-mediated data.

Data papers. A data paper is a peer-reviewed article whose primary purpose is to present a dataset through a narrative. The value of data papers is at least threefold: they increase the usefulness of the dataset by functioning as extended metadata, they are a mechanism by which researchers can receive acknowledgement for published data, and they serve as an entry point to biodiversity data standards and to GBIF (Chavan and Penev 2011, Penev et al. 2017). In this way, data papers both enable and incentivize data publication. This represents a cultural shift in biodiversity research, marking a turning point from an earlier status quo where lack of incentivization was seen as a barrier to data publication (Penev et al. 2017, Groom et al. 2020). GBIF provides a range of tools to support data paper production, and there are now at least 40 peer-reviewed journals with the capacity to publish datasets from data papers directly to GBIF. These journals play an especially important role in facilitating data publication by researchers who are not affiliated with an endorsed data publishing institution. At the time of writing, over 700 data papers have published data to GBIF, including through a number of sponsored calls resulting in special issues and thematic collections.

Data extracted from literature. A growing source of GBIF-mediated data is from the published literature beyond data papers. A GBIF publisher and organizational node, Plazi, has developed an automated pipeline to extract taxonomic or occurrence data from the literature and standardize it for publication (Agosti et al. 2023). Plazi has processed nearly 50% of the datasets published in GBIF that provide material citation records and is a critical source of information on newly published species.

Machine-derived data. New technological advances were catalysts for each of the major phases in expanding GBIF-mediated data and continue to drive development of GBIF workflows. New data types are increasingly enabled by digital mediation, such as machine vision for taxonomic identification (as with participatory science photography, remotely sensed imagery, or digitized specimen imagery); automated monitoring schemes, including camera traps and passive acoustic monitoring; and high-volume data inputs that require automated processing, as in many molecular methods (Steinke et al. 2025). Alongside the positive impacts of automation and digitization on many GBIF-mediated data and processes, it should be emphasized that all GBIF-mediated datasets require human accountability. A cornerstone of GBIF’s approach to data quality is the association of all data with points of contact who can respond to queries and manage quality control.

A query of GBIF-mediated data integrates individual records from discrete datasets while retaining dataset-level metadata and provenance. Datasets are published as one of four dataset classes: occurrence, sampling event, checklist, or metadata-only. These classes reflect a range of typical dataset complexity and richness scenarios (though in practice dataset class may be limited by data publishers’ resources to invest in data publishing rather than inherent dataset complexity).

Occurrence datasets—records of organisms at specific times and locations—have been the original focal point for GBIF, and were by far the most prevalent class of datasets mediated by GBIF for its first twenty-five years. Around 2016 it became possible to publish more complex biodiversity data with sampling event datasets. Sampling event datasets invert the paradigm associated with occurrence datasets, emphasizing the sampling event—the survey design and sampling effort, including temporal and spatial effort and taxonomic targets—and clustering occurrence records within the context of the event. Documenting and quantifying the survey enables inference of taxon absence and provides interpretational opportunity for taxon abundance. Such data are native for ecologists and for the biological monitoring community. This was a major milestone for GBIF, made possible by years of collaborative work at TDWG to develop the Darwin Core Event Class and further improved with the introduction of additional terms ratified by TDWG in 2024 as the Humboldt Extension (Guralnick et al. 2018, Sica et al. 2026).

Although sampling event data currently make up a relatively small proportion of the total data mediated by GBIF, the transition to a much stronger integration of sampling event data is expected to be a central focal point for GBIF in the years to come, responding to policy demands for monitoring of biodiversity. Support for sampling event data is facilitated by event-centric services in GBIF, most notably systematic infrastructure improvements to support the latest Darwin Core data model, the Data Packages (DwC-DP; Wieczorek et al. 2025). The DwC-DP data model was designed to more realistically reflect the natural complexity of ecological data, accommodating a variety of event types, varied sampling protocols, and multiple data types. GBIF will incrementally implement infrastructural support for DwC-DP over the next few years.

In addition to the occurrence data that have driven the first twenty-five years of GBIF and the sampling event data that are now coming to prominence, GBIF recognizes two dataset classes that are simpler in structure. The simplest, metadata datasets, consist only of the information that a dataset exists but is not yet published through GBIF. Metadata-only datasets are important for inventories and benchmarking, helping set priorities for future data collection and digitization.

Finally, checklist datasets report a list of taxa that has been created for a specific context (for instance, flora of a protected area, invasive or protected species in a region, or a taxonomic revision). In other words, the name in a checklist carries the central function, in contrast with spatiotemporal evidence that holds the central role in occurrence datasets. Well-curated checklists of Linnaean names and reference libraries of molecular taxa play a key role in GBIF’s indexing abilities, as any recording in the incoming occurrence or sampling event dataset is matched against the centrally assembled backbone of taxon names and operational taxonomic units.

The taxonomic backbone enables the indexing of GBIF-mediated data according to standardized taxonomies. After developing and maintaining the widely used GBIF taxonomic backbone, in 2026 GBIF has adopted the Catalogue of Life as its taxonomic source (Bánki et al. 2026). This transition was the result of a multiple year collaboration between GBIF and Catalogue of Life in developing a joint infrastructure, ChecklistBank. The infrastructure contains tools to compare among taxonomic checklists and to build new taxonomic lists from sources such as ITIS, Plazi, World Flora Online, World Register of Marine Species, The Paleobiology Database, and Index Fungorum. Catalogue of Life uses ChecklistBank to build its monthly taxonomy, guided by its community of taxonomic experts. This enables GBIF to contextualize and integrate many different taxonomies as they evolve.

Not all biodiversity data shared with GBIF contain a traditional Linnean classification. GBIF now incorporates DNA based classifications, such as iBOL, Barcode Index Numbers, and UNITE’s fungal species hypotheses in the COL taxonomic backbone. In 2026, GBIF embarked on a program of managing DNA reference libraries published to GBIF with a system to reannotate the identifications of DNA-derived occurrences to the ever-improving DNA reference libraries published to GBIF. These species hypotheses, which are incorporated into the monthly releases of the Catalogue of Life taxonomy with the Linnean species, now number several million taxa and provide taxonomic reference for hundreds of millions of GBIF occurrences.

Interoperability across GBIF-mediated data is enabled by data standards that provide a consistent, replicable structure for data formatting. A successful data standard accurately reflects relationships among multiple characteristics of the data; it is readable by both humans and machines; and it is flexible enough to accommodate a sufficient variety of data (Blum and Bloom 2025). And in a network that is all about free and open access, perhaps most crucially, it is community-driven and obtains broad consensus and adoption within the relevant communities of data publishers. From the start, GBIF was built on community-developed open data standards and maintained a close relationship with TDWG and the data standards community (Robertson et al. 2014, Blum and Bloom 2025). In this collaborative environment, GBIF quickly emerged as one of the most common places—and in many cases likely the only place—that many researchers and users of biodiversity data encounter the concept of data standards, which has effectively placed GBIF in the position of normalizing data standards among the broader biodiversity data community. Thus, GBIF plays a significant role in the testing, maintenance, and development of standards.

There are a number of data standards that are central to GBIF’s historical development and present operations. These are:

Darwin Core Standard. Darwin Core is the primary standard for occurrence data, and the main standard by which GBIF-mediated data have been structured (Darwin Core Task Group 2009, Wieczorek et al. 2012). Darwin Core data are published in the Darwin Core Archive (DwC-A) file format, which consists of a ZIP file containing interconnected text files that hold the core information and any extensions. Beyond the mandatory spatiotemporal and taxonomic information, most occurrence data include richer information, standardized with optional data fields and extension files. The introduction of new and revised extension files is therefore a key way that further development of Darwin Core takes place. Of particular importance to GBIF, the Event core extension has since 2015 allowed structured survey data to be standardized as DwC-A files, and the recent Humboldt Extension greatly extends this capability (Guralnick et al. 2018, Sica et al. 2026). Most recently, the Darwin Core Data Package (DwC-DP) was introduced to enable more flexible and relational application of standard Darwin Core vocabulary, accommodating the structures of the Frictionless Data Package framework (Fowler et al. 2017, Wieczorek et al. 2025). The DwC-DP structure will exist in parallel with DwC-A, with increasing infrastructure support for DwC-DP planned in coming years.

Access to Biological Collections Data Schema (ABCD). ABCD is an analogue to Darwin Core that originated with the Biological Collection Access Service for Europe (BioCASe) and is used by many natural history museums, botanical gardens, and zoological institutes (Access to Biological Collections Data Task Group 2005). While Darwin Core is the primary and preferred data standard for occurrence data mediated by GBIF, GBIF has developed workflows to translate ABCD datasets to Darwin Core format.

Ecological Metadata Language (EML). All datasets on GBIF are accompanied by metadata files containing additional information about the data. These use the Ecological Metadata Language (EML), which is a metadata standard. EML (Jones et al. 2019, v2.2.0) is a community-maintained, XML-based metadata specification that defines a modular vocabulary for documenting ecological and environmental datasets, covering identity, methods, and taxonomic, geographic, and temporal coverage.

The quality of GBIF-mediated data relies, first and foremost, on the quality of data contributed by each data publisher. An analogy to a library is fitting: GBIF sets standards for data quality, provides guidance to data publishers, and maintains oversight through node endorsement of data publishers. Yet, much as it is outside the scope of a library to edit each book on its shelves, it is infeasible and outside the scope of GBIF to perform detailed quality control on GBIF-mediated data. Therefore, GBIF requires data publishers to be diligent about quality control, and asks users to report any quality concerns to GBIF or the data publisher.

To help data users conceptually navigate the new capacity for data integration and synthesis offered by GBIF, GBIF advanced a framework of data quality not in absolute terms but in terms of contextual fitness-for-use (Chapman 2005a, Chapman 2005b). This remains an essential framework for applying GBIF-mediated data. To assist users in evaluating the fitness of data for potential uses, GBIF flags several dozen common data quality issues during data processing. These include geospatial issues, taxonomic issues, date issues, and errors in vocabulary for standardized data fields, among others. Affected data are flagged, automatic interpretations of data quality issues are applied in certain cases, and the issues are documented on the dataset page of the GBIF portal so that data users can decide how to further clean the data for their intended use. To support data publishers in publishing high quality data, GBIF offers a data validator service that produces a report flagging potential issues so that the publisher can resolve them before publishing the dataset. Because carrying out quality control is often reported to be time-consuming for both data publishers and data users, GBIF aims to continue this work area and promotes the use of hosted portals that allow for more tailored data filtering.

A broader issue related to fitness-for-use is the prevalence of geographic and taxonomic asymmetries in data coverage (Heberling et al. 2021, Mandeville et al. 2023, Steinke et al. 2025). These asymmetries largely echo those that have plagued biodiversity research since its beginnings, including underrepresentation of highly biodiverse areas, especially in the Global South; overrepresentation of charismatic species, particularly birds; and underrepresentation of taxa that are less accessible to observe and identify, a category that includes many marine taxa as well as many hyperdiverse, cryptic, and undescribed taxa (Troudet et al. 2017, Hughes et al. 2021). The need to overcome these asymmetries in global research was a major factor in the initial call for a global biodiversity information facility, and continuing to overcome them remains a central challenge for GBIF. GBIF’s efforts to expand Participation across the globe, described in detail in the “GBIF capacity development” section, aim to play a role in filling geographic data gaps, and the integration with Catalogue of Life and growing services to support publishing of novel data types including molecular and remotely sensed data will contribute to filling taxonomic data gaps. Still, despite the role GBIF is poised to play in addressing these asymmetries, they are larger issues than GBIF can resolve alone; continued progress will rely on broader global shifts in scientific data culture, funding, and capacity.

There are several pathways for publishing data via GBIF, which are regularly updated to accommodate new types of data and publishers. To date, the majority of GBIF-mediated data are published using the GBIF Integrated Publishing Toolkit (IPT; Robertson et al. 2014). Other data sharing protocols (DiGIR, BioCASE, TAPIR) were also important in the early days of GBIF. The IPT, which is the typical publishing pathway today, is a free, open-source software developed by GBIF. Its core function is to create Darwin Core files, register them with GBIF, and make them accessible online. Its capacity to create and manage repositories of data means that it also functions as a national repository to enable countries to meet national data hosting requirements.

In many cases, data publishing institutions access the IPT via an IPT installation managed by the Participant node with which they are affiliated. Data publishing institutions can opt to install and run their own individual instances of the IPT, but this requires a relatively high level of technical resources and expertise as well as an ongoing update and backup routine, so it is not feasible for all data publishers. Because many data publishers are neither equipped to run an IPT installation nor affiliated with a Participant node, several nodes also function as data hosting centers for publishers outside their own country. In cases where none of the previous options are possible, the GBIF Secretariat can establish publisher accounts on hosted IPT instances that rely on shared hardware, software, and storage services, offering publishers basic backups and software updates in a format they can migrate later to another IPT installation.

The third major version of the IPT was released in 2024, including new support for Frictionless Data Packages (Fowler et al. 2017, Bubnicki et al. 2024). In the same year, GBIF launched the pilot phase of the Metabarcoding Data Toolkit as part of an ongoing effort to improve GBIF’s integration of DNA metabarcoding data (GBIF Secretariat 2024b). As high-throughput data derived from large automated monitoring programs (e.g., eDNA, camera trap, acoustic monitoring) make up an increasing proportion of GBIF-mediated data, API linkages for data publishing are expected to become more important. Nodes and the broader GBIF community support further development of data publishing infrastructure; for instance, nodes have developed publishing solutions independent of the IPT, and community-developed tools have on occasion been developed to lower the technical threshold for new publishers and extend publishing options to new research communities.

There are many ways for users to query, view, and download GBIF-mediated data, increasing in complexity from very simple to highly customizable. The simplest approach is to use a GBIF portal; all GBIF-mediated data can be searched and freely accessed via the GBIF.org portal and the GBIF-supported hosted portals and Living Atlases maintained by Participant nodes. Using the web portals requires no programming experience and allows for visually supported searching, filtering, and downloading of data, as well as giving access to a range of additional tools and metrics. GBIF-mediated data can be downloaded in multiple formats to meet different user needs, and technical advice is provided through training modules, documentation, and blog articles to help users work with the data. Thanks to the efforts of network’s volunteer translators, GBIF.org appears in twelve languages, including the six official UN languages: English, Arabic, Chinese (simplified and traditional), French, Russian, Spanish, Czech, Japanese, Polish, Portuguese, and Ukrainian.

A more flexible, but somewhat more technically demanding approach to access data is via the GBIF API and associated API services. API services allow continued access to GBIF-mediated data through other systems and can be the basis for the development of tools that interact with the data. Examples include packages to access the API within common coding environments, such as rgbif for R and pygbif for Python (Ancona et al. 2007, Chamberlain and Boettiger 2017, Chamberlain et al. 2026). The API allows a user, among other things, to make advanced queries, include these queries in scripts and workflows for automated use, and embed GBIF-mediated data into other websites. Finally, GBIF-mediated data can be accessed via cloud computing services; every month GBIF takes a snapshot of occurrence information and uploads it to three cloud repositories. Technical guidance is provided for querying and accessing cloud-based data via Google BigQuery and other services. As with data publication, the services available for data query and download are updated regularly.

User support is available to help users access data, facilitate peer support (e.g., via the GBIF Community Forum), and enable feedback on user needs. Nodes are the primary point of contact, but there is also a central helpdesk whose service is distributed across the network of nodes, supported by regional support teams. Regardless of the access method, downloading GBIF-mediated data requires users to be logged in with a free GBIF account, to accept the GBIF Data User Agreement, and to agree to cite data following best practices (GBIF 2017). Good citation practices, including citing the DOI of datasets and data downloads, support scientific transparency and reproducibility, deliver credit to data originators and data publishers, and reinforce the value and demonstrate the impact of open data.

Thorough technical documentation of the GBIF infrastructure is beyond the scope of this article. Instead, we provide a concise high-level overview of the architecture and infrastructure underlying GBIF and refer interested readers to the more detailed technical documentation available through the GBIF portal (GBIF Secretariat 2026b).

The GBIF infrastructure is globally distributed and comprises several interconnected layers. Following data publication, published datasets are registered through the GBIF Registry, a central catalogue maintaining metadata on publishers, institutions, datasets, and the technical endpoints necessary to access them. GBIF’s harvesting and processing infrastructure continuously retrieves registered datasets and integrates them into a centralized data platform. Taxonomic interpretation, supported by ChecklistBank, enables the reconciliation of scientific names across datasets (Bánki et al. 2026). Original published records are preserved alongside a series of standardized interpretation and enrichment processes, including taxonomic matching, geographic validation, and temporal normalization, resulting in searchable and interoperable data products.

Processed information is stored in a scalable data warehouse following a medallion architecture, in which successive layers preserve original records while generating increasingly standardized and enriched representations. This platform is implemented using open source distributed big-data technologies, including Apache Hadoop and Apache Spark, enabling the processing of billions of biodiversity records and generation of derived products. The resulting data products are exposed through dedicated search indexes based on Elasticsearch, providing real-time querying capabilities across occurrences, species, datasets, and institutions, and through asynchronous data export APIs built on SQL standards. These are accessed through open RESTful APIs that provide machine-readable access to GBIF data and power the GBIF web portal, hosted portals, R and Python client libraries, and external applications built on top of GBIF.

The GBIF codebase is developed openly and released under open licenses, benefiting from an active community of developers and contributors who provide guidance through issue reporting, translations, and code contributions (e.g., hosted portals and collaborations between the GBIF and Living Atlas software platforms). Researchers and developers wishing to contribute to GBIF’s data processing infrastructure can consult the developer contribution guidelines for the GBIF pipelines project (GBIF Secretariat 2026a).

In addition to its core services, GBIF leverages its informatics infrastructure to capture a range of metrics related to its performance and impact. Impact tracking is an essential way that GBIF can learn more about the professional communities that engage with its services and identify opportunities to expand those services. Similarly, impact tracking enables the GBIF community to learn and take inspiration from each other. It is also an important means of demonstrating the return on investment for the national governments that fund GBIF’s operations and the individuals who volunteer their time for the GBIF network, and is essential for conveying the value of GBIF to nations that are not yet Participants.

Citation tracking is the primary way of tracking GBIF’s impact on scientific research and policy. The GBIF Secretariat maintains a citation tracking program that captures scientific manuscripts and policy reports that follow best practices for data citation with a DOI as well as (to the extent possible) manuscripts that have not cited their use of GBIF-mediated data directly. Use of DOI-based data citations increased from 9% in 2017 to ~50% in 2026. All articles using GBIF-mediated data are openly searchable within the same portal infrastructure as GBIF biodiversity data with direct links between articles and the data they use. Many data uses are summarized in brief reports in the GBIF portal, and GBIF’s annual Science Review publication gathers key research applications. Other impact-related reports, including from projects funded under capacity development schemes, are also indexed in the GBIF portal.

At its core, the GBIF community is made of individuals. Not only do individuals use and publish GBIF data, serve on staff in data publishing institutions and nodes, and contribute directly to the GBIF network as volunteers, but they also act as champions of GBIF who advocate for free and open biodiversity data, initiate connections between their institutions and GBIF, and call for the establishment of new GBIF Participants. GBIF’s strategy for capacity development includes the growth of a community of practice that empowers all individuals in the GBIF community to contribute to its mission.

Participant nodes are a first point of contact for supporting individuals and institutions. This direct support is supplemented by a range of other resources including quarterly Data Use Club sessions, the GBIF Community Forum, and a helpdesk email address. Individuals and institutions can also access GBIF training resources, which follow digital-first practices to enable transparency and accessibility. GBIF recognizes outstanding contributions of the network through annual GBIF awards. To inspire, incentivize, and give visibility to individual GBIF community members who make especially innovative contributions to data mobilization and use, the Graduate Researchers Award highlights innovative research by graduate students in Participant countries and the Ebbe Nielsen Challenge recognizes innovative applications of GBIF-mediated data.

The capacity of individuals and institutions to engage with GBIF is strengthened by an overall increase in global data literacy and support for open science. Though a deep focus on this form of global capacity development is beyond the core scope of GBIF, GBIF supports its network in promoting such activities (for example through partnership with DataCamp Donates) and at times participates directly in strategically relevant capacity building on broader themes related to its mission.

GBIF Participant nodes carry out the primary work of GBIF, so developing capacity at the node level is critical. Nodes vary greatly across the GBIF network, with variation in hosting institutions, scales of operation, and budgets, so node capacity development must be customizable. Although training materials are available for nodes, especially for onboarding of new nodes, knowledge exchange among nodes tends to be the best resource for ongoing capacity development. The main forum for this is the Nodes Committee, one of GBIF’s three standing committees. Since 2019, the Nodes Committee has also identified annual strategic objectives for the nodes via a Nodes Implementation Plan. GBIF encourages regular monitoring of node capacity to support strategic planning and offers an online survey tool that helps nodes assess their capacity.

Nodes can access a limited amount of financial support to build capacity in specific projects. Although the main financing for nodes is established and allocated by the Participant, GBIF offers small grants to fund regional capacity development projects through the Capacity Enhancement Support Programme. These small grants often allow nodes to take on projects to address specific regional needs that might not be funded within their ordinary budgets, and are especially important in regions with lower financial capacity. To further amplify node innovations, GBIF introduced a non-monetary biennial node award in 2023.

The GBIF hosted portal service is an increasingly important offering for supporting nodes in engaging their national and regional communities. Hosted portals display a subset of GBIF data relevant to a particular community. Nodes can apply to establish a national hosted portal, which they can customize with relevant languages and visual identity. This allows nodes to more easily promote data mobilization and the use of GBIF-mediated data within their communities, while generating significant cost savings for the node. Hosted portals can also be developed to cover thematic areas, so nodes can offer hosted portals as a service to data publishers in their networks. In some cases, nodes take further steps to develop infrastructure that makes data accessible in ways that meet particular needs of their node. The Living Atlases and Symbiota Support Hubs are examples. These are open source platforms that not only provide national data portals but also act as hubs for technical development and data sharing.

Regional-level organization is an important part of GBIF’s global mission. Regional capacity development enables stronger nodes to support less experienced ones in their region, and nodes can build upon physical proximity (and in some cases shared cultural, governance, or language contexts) to collaborate and act collectively on regional issues. Regional capacity development is also a necessary starting point for promoting biodiversity data mobilization and the use of GBIF-mediated data in countries that are not yet GBIF Participants, which can eventually galvanize stronger national engagement and Participation. In addition to representation in the Nodes Committee, each node is affiliated with one of six regional subcommittees, which generally meet more frequently and often work together with organizational Participants and external partners at the regional level.

Since the early 2020s, GBIF has intensified its efforts to serve geographic areas that are not yet supported by nodes. Existing nodes are important leaders in these efforts, alongside growing contribution from contracted regional support teams. Regional support teams have become a key way that GBIF serves areas without direct node support. Their work includes engaging with research institutions, supporting data publishing, capacity development, supporting new nodes, and providing GBIF with information on regional capacity needs.

Critical financial support for regional capacity development has been provided through the Biodiversity Information for Development (BID) program, a multiyear program primarily funded by the European Union and led by GBIF. The BID program funds projects that develop capacity for data mobilization and use in Africa, Latin America and the Caribbean, and the Pacific islands, with a particular focus on data that address national and regional needs. Activities funded by BID have been important for strengthening national and regional biodiversity information facilities in the target regions.

At the global level, GBIF leads by example in and beyond the biodiversity domain promoting free and open access to data, a culture of data transparency and data citation, and the use of data standards.

GBIF has become a fundamental data infrastructure underpinning biodiversity-related science, a vast category that spans fields from environmental research to human health to agriculture and forestry, among many others. By facilitating free and open access to data, GBIF lowers global technological and resource barriers to research, enables data interoperability, and extends the longevity of data by enabling its reuse. Yet GBIF’s impact on scientific research is still greater than the sum of its individual datasets; GBIF adds further value through the integration of each dataset with the GBIF infrastructure, such that its impact increases nonlinearly as the GBIF infrastructure grows (Heberling et al. 2021, Deloitte 2023, Steinke et al. 2025).

The impact of GBIF is clearly seen in the role of GBIF-mediated data in driving scientific research. The body of 15,000 peer-reviewed research articles that have cited GBIF-mediated data continues to grow at a rate of more than eight new articles per day. The extent of data indexed by GBIF increasingly gives rise to new forms of analysis and new research questions that would be infeasible or impossible without such a global infrastructure, with nearly half (47%) of GBIF users in a 2022 user survey reporting that their research would have been impossible without GBIF (Deloitte 2023). Recent systematic reviews reveal a wide and growing range of methodologies used to leverage the unique potential offered by GBIF-mediated data (Heberling et al. 2021, Steinke et al. 2025).

The impact of GBIF on science can be further measured in the time savings and efficiencies it facilitates; in addition to the users whose research would be impossible without GBIF, a further 41% reported that it could only have been possible with significantly greater time and effort (Deloitte 2023). Such results demonstrate the evolution of GBIF's usefulness over time, showing substantial improvement over earlier assessments (e.g., Faith et al. 2013). Today it has been clearly demonstrated that GBIF-mediated data and services greatly reduce the cost, labor, and technical difficulty of research compared to the effort that would otherwise be required for individuals, institutions, or nations to replicate aspects of the GBIF infrastructure. This saves time and resources that can be redirected to more critical research needs. The annual value of researcher time saved by GBIF was estimated in 2021 to be €35 million (Deloitte 2023).

Through provision of free and open access to data and informatics services, GBIF expands the research capacity of countries that would otherwise be limited by technological or resource constraints. The scientific impact of GBIF’s capacity development in resource constrained regions is seen in the impacts of its BID program: in the first eight years of the program, 703 datasets were mobilized from the target regions, which were downloaded 1.8 million times, resulting in 715 publications and 34 IUCN Red List assessments (CSIL 2023). In total GBIF has funded over 102 BID projects and 61 Capacity Enhancement Support Programme projects as of 2026 (CSIL 2023). The signature of the GBIF Memorandum of Understanding by 15 countries and two organizations involved in activities organized as part of the BID program is further illustration that capacity development efforts generate genuine progress in increasing scientific capacity in countries facing resource constraints.

GBIF facilitates and streamlines the application of biodiversity data to evidence-based policy decision-making and impact assessment. Major international policy frameworks underscore the necessity of free and open biodiversity data and increasingly identify GBIF explicitly as indispensable to meeting international policy targets. Among others, GBIF is endorsed by the Convention on Biological Diversity, to which it is an official Observer Organization; the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) relies on GBIF-mediated data for many of its assessments; the International Union for Conservation of Nature (IUCN) indicates that GBIF-mediated data underpin 25% of its Red List species assessments; 92% of GBIF users indicate that their research is applicable to the UN Sustainable Development Goals; and GBIF is recognized as a key partner of the Global Environmental Facility (Deloitte 2023). GBIF’s relationships with many entities at the science-policy interface position it as a critical facilitation point between the research and policy communities.

In addition to its direct engagement with international and intergovernmental bodies at the science-policy interface, much of GBIF’s impact on policy comes at the national and regional level, with GBIF nodes playing instrumental roles in national assessments and supporting compliance with national and international policy commitments. GBIF infrastructure enables government agencies to meet commitments in areas including open science, environmental impact assessment, and evidence-based decision-making on issues like land use planning, climate resilience planning, and public health, among others (Groom et al. 2025).

GBIF first gained a recognized role in biodiversity policy as Parties to the Convention on Biological Diversity reported on the growth of GBIF-mediated data as an indicator for the Aichi Targets. Under the current Kunming-Montreal Global Biodiversity Framework, GBIF-mediated data are used across multiple indicators, including the headline indicator for Target 6 on invasive and alien species, in Target 4 for species conservation, in Target 20 for capacity mobilization, and in Target 21 to ensure access to best available data (Convention on Biological Diversity 2022b). Use of GBIF-mediated data is well represented in the 7th National Reports to the CBD, in which nations reported on progress towards their National Biodiversity Strategies and Action Plans in connection to the Global Biodiversity Framework.

The Global Biodiversity Framework also includes Target 15, in which nations assess and disclose biodiversity risks and impacts by the private sector (Convention on Biological Diversity 2022b). As of the end of 2025 nearly 150 private sector businesses have contributed over 14 million occurrences to GBIF, mainly from impact assessment data. Many companies support evidence-based decision-making and inform corporate reporting by making use of free and open location-specific data and information mediated by GBIF. In addition to national biodiversity reporting, the private sector has self-organized around corporate reporting on its biodiversity impact through the Taskforce on Nature-related Financial Disclosures (TNFD). Following from and improving on the example from climate reporting, the TNFD sets recommendations and guidance on their nature-related dependencies, impacts, risks, and opportunities. The TNFD explicitly calls on businesses to publish biodiversity-related data via GBIF in Recommendation 5, “Incentivise corporate nature data exchange” (Taskforce on Nature-related Financial Disclosures 2025).