{"paper_id":"2404.17369","title":"Assessing the Potential of AI for Spatially Sensitive Nature-Related Financial Risks","authors":["Steven Reece","Emma O'Donnell","Felicia Liu","Joanna Wolstenholme","Frida Arriaga","Giacomo Ascenzi","Richard Pywell"],"paper_date":"2024-04-26","paper_type":"perspective-policy","paper_url":"https://arxiv.org/pdf/2404.17369v1","page_url":"https://thequant.space/flowcharts/assessing-the-potential-of-ai-for-spatially-sensitive-nature/","scores":{"empirical_rigor":4,"math_complexity":3,"quadrant":"Philosophers"},"proposes":{"status":"available","statement":"The paper proposes AI-based modelling approaches for two use cases – Brazil beef supply chains and UK water utility catchments – to attribute nature-related impacts and inform sustainable finance decisions.","quote":"This report presents potential AI solutions for models of two distinct use cases, the Brazil Beef Supply Use Case and the Water Utility Use Case.","summary":"There is growing recognition among financial institutions, financial regulators and policy makers of the importance of addressing nature-related risks and opportunities. Evaluating and assessing nature-related risks for financial institutions is challenging due to the large volume of heterogeneous…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The farm state is independent of the specific sourcing abattoir and the probability of Forest Code compliance depends only on the farm state.","type":"model","stated":"explicit","quote":"Note, in the equations above, we have assumed that the farm state is independent of the specific sourcing abattoir and that the probability of farm Forest Code compliance, or the NRP impact value 𝑁, depends only on the farm state and not, for example, on the sourcing beef supplier.","support":"full"},{"assumption":"Conditional independence allows the joint probability of farm, supplier, logistics and compliance variables to be factored into separate terms.","type":"statistical","stated":"explicit","quote":"The joint 28 probability can be factored by assuming conditional independence.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Cadastro Ambiental Rural (CAR)","provider":"Brazilian federal and state governments","kind":"alternative","access":"public","period":null,"universe":"properties/farms in Brazil","quote":"Survey of relevant environmental datasets and methodologies Data Source Description Project Relevance Cadastro Ambiental Rural (translates as ‘Environment al Rural Registry’) Brazil's Forest Code established a mandatory and georeferenced registry of farms aimed at advancing the implementation of its natural ecosystem protection and restoration requirements. Access to >6m properties (over 544m ha), which provides us","support":"full"},{"name":"World Database on Protected Areas (WDPA)","provider":"UNEP‑WCMC / IUCN","kind":"alternative","access":"public","period":null,"universe":"global protected areas","quote":"World Database on Protected Areas (WDPA) The World Database on Protected Areas (WDPA) is a joint project between UN Environment Programme and the International Union for Conservation of Nature (IUCN), managed by UNEP-WCMC. The WDPA uses the IUCN definition of a protected area as the main criteria for entries included in the database.","support":"full"},{"name":"Key Biodiversity Areas (KBA)","provider":"IUCN","kind":"alternative","access":"public","period":null,"universe":"global biodiversity hotspots","quote":"Key Biodiversity Areas (KBA) KBAs represent the most important sites for biodiversity conservation worldwide, and are identified nationally using a Global Standard from the International Union for the Conservation of Nature (IUCN). Key Biodiversity Area data is required for the ‘significance’ element of the NRP methodology.","support":"full"},{"name":"Species Threat Abatement and Restoration Metric (STAR)","provider":"IUCN Red List via IBAT","kind":"alternative","access":"public","period":null,"universe":"global species threat data","quote":"20 Species Threat Abatement and Restoration Metric (STAR) The Species Threat Abatement and Restoration Metric (STAR) allows quantification of the potential contributions that species threat abatement and restoration activities offer towards reducing extinction risk across the world. STAR is calculated from data on the distribution, threats, and extinction risk of threatened species derived from the IUCN Red List of","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"none"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"none","findings":[{"finding":"AI modelling can attribute biodiversity loss and gain to each actor in the value chain, enabling investors to see specific impacts and mitigate greenwashing.","kind":"theoretical","quote":"Our asset level modelling approach, with the help of AI, exploits a wide range of detailed local heterogenous data that can attribute biodiversity loss and gain to each actor in the value chain.","support":"full"},{"finding":"Uncertainty is explicitly modelled, guiding further data gathering and model refinement.","kind":"theoretical","quote":"Uncertainty is modelled explicitly and this both informs and directs further efforts for data gathering and model refinement.","support":"full"},{"finding":"AI techniques require substantially more computing power than traditional methods, posing practical challenges.","kind":"negative","quote":"Practically, AI techniques typically require much more computing power than more traditional methods.","support":"full"},{"finding":"Asset‑level data granularity is a limitation because companies often lack or are unwilling to disclose such detailed data.","kind":"negative","quote":"Furthermore, our approach requires asset-level data to achieve the necessary granularity, which has proven challenging as companies either do not have data to the required granularity or they are unwilling to disclose it voluntarily.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"not_extracted","supports":[],"contradicts":[],"qualifies":[],"compared_with":0,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":2,"extracted_at":"2026-11-14T19:57:42Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}