EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Is conservation AI ready? Quantifying the challenges of automated species identification in Colombia

Gustavo Nicolas Paez Salamanca iD

Biodiversity Data Journal · 05.11.2026 · 14

DOI: 10.3897/bdj.14.e208137 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Biodiversity Data Journal, online erschienen am 05.11.2026. Verfasst von Gustavo Nicolas Paez Salamanca. Beteiligte Einrichtung: Data and Analytics Manager, UNICEF Asia Pacific Regional Office, Bangkok, Thailand Data and Analytics Manager, UNICEF Asia Pacific Regional Office Ba…. Die Arbeit nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Multimodal artificial intelligence (AI) is increasingly adopted by conservation organizations to overcome taxonomic bottlenecks. However, its reliability may systematically fail for sparsely documented and recently described taxa. This study evaluates automated species identification using 10,000 research-grade iNaturalist images of Colombian vertebrates across amphibians, birds, fishes, mammals, and reptiles. Testing OpenAI GPT-5.2 with localized prompts but no taxonomic metadata, we found that species-level accuracy scales non-linearly with global photographic representation and decays severely for recently recognized taxa. At the 10th percentile of data availability, predicted accuracy collapsed to 6–12% for most groups, with mammals performing best at 41%. Furthermore, severe taxonomic hallucinations frequently crossed class boundaries, particularly among fishes and reptiles. These findings show that AI errors are not randomly distributed but are strongly associated with public photographic representation and taxonomic recency. While general-purpose AI offers immense value for rapid sorting, its operational integration requires strict safeguards, including forced abstention protocols, geographic filters, and mandatory expert escalation for rare or recently described species.

Schlagwörter

Stichwörter der Autor:innen

artificial intelligencebiodiversity monitoringcitizen scienceiNaturalistspecies identificationtaxonomic bias

Autor:innen und Institutionen

  • Data and Analytics Manager, UNICEF Asia Pacific Regional Office, Bangkok, Thailand Data and Analytics Manager, UNICEF Asia Pacific Regional Office Bangkok Thailand https://ror.org/03xs42a74.

ORCID-Profile: Gustavo Nicolas Paez Salamanca

Zitieren

APA

Paez Salamanca, G. N. (2026). Is conservation AI ready? Quantifying the challenges of automated species identification in Colombia. Biodiversity Data Journal, 14. https://doi.org/10.3897/bdj.14.e208137

Vancouver

Paez Salamanca GN. Is conservation AI ready? Quantifying the challenges of automated species identification in Colombia. Biodiversity Data Journal. 2026;14. doi:10.3897/bdj.14.e208137

BibTeX
@article{PaezSalamanca2026,
  author = {Paez Salamanca, Gustavo Nicolas},
  title = {Is conservation AI ready? Quantifying the challenges of automated species identification in Colombia},
  journal = {Biodiversity Data Journal},
  year = {2026},
  volume = {14},
  doi = {10.3897/bdj.14.e208137}
}
Kostenlos starten →