---
title: "share the finding, protect the place"
canonical_url: https://ensurance.app/guide/share-the-finding-protect-the-place
markdown_url: https://ensurance.app/guide/share-the-finding-protect-the-place.md
subtitle: practical choices for sharing useful evidence while protecting sensitive information
category: how-to
---

# share the finding, protect the place

*practical choices for sharing useful evidence while protecting sensitive information*

An owl recording can be useful evidence and a very precise set of directions. Sharing it well means deciding which information each recipient needs.

Privacy-preserving ecological evidence keeps a useful claim available while limiting disclosure of sensitive source material. That can mean publishing a broader location, granting access to an authorized reviewer, or proving a narrowly defined statement without exposing its inputs. The right choice depends on the question, the source's terms, and the harm disclosure could cause.

Start with the question. A researcher estimating regional distribution, a steward making a site decision, and a reader learning about habitat may need different views of the same record.

## one recording, several possible views

Consider a hypothetical field team monitoring northern spotted owls. It has an acoustic file, a timestamp, a precise recorder location, and a reviewed identification. This is an illustrative record; no actual owl location is supplied here.

Northern spotted owls inhabit structurally complex forests across parts of the Pacific Northwest. The U.S. Fish and Wildlife Service's species account provides general ecological context without requiring us to publish a monitoring site's coordinates. [USFWS: Northern Spotted Owl](https://www.fws.gov/species/northern-spotted-owl-strix-occidentalis-caurina).

The team wants its work to inform habitat care. It could make a public summary available while allowing a qualified reviewer to inspect the underlying recording under agreed conditions. Those views answer different questions. The summary helps explain a regional concern; the restricted file allows closer examination of the detection and its context.

The team must also inspect the material around the obvious sensitive field. A filename, a spoken place name in the recording, or a link to another record could reveal the location after coordinates are removed. Treat the disclosure as a whole.

## choose the least detail that answers the question

GBIF's guidance on sensitive species occurrence data describes balancing useful access with the risk of harm. It recommends documenting sensitivity and any generalization, keeping the source record intact, and reviewing restrictions. A coarser public location should carry its actual precision so a reader does not mistake it for a survey point. [GBIF: Generalizing Sensitive Species Occurrence Data](https://docs.gbif.org/sensitive-species-best-practices/master/en/).

Use that principle to make a disclosure choice:

| option | what the recipient receives | when it can help |
|---|---|---|
| Public summary | An approved conclusion, broad location or time window, method, and stated limits. | Education or analysis that can use the reduced detail. |
| Restricted source record | Detailed material available to named, authorized recipients under agreed uses. | Identification checks, site planning, or an independent review that needs the original. |
| Proof of a defined statement | Verification of a particular relationship over hidden inputs. | A decision that genuinely needs the statement rather than the underlying details. |

Write down why that view is sufficient. If generalization removes information an analysis needs, say so. Access control can be more useful than a public summary for that specific task. A proof can be more complicated than either, and should earn that complexity.

## what a zero-knowledge proof could establish

A zero-knowledge proof, or ZK proof, can establish a mathematical statement without revealing additional information used to demonstrate it. That is the scope described by NIST's work on privacy-enhancing cryptography. Its usefulness depends on stating the relationship precisely. [NIST: Zero-Knowledge Proofs](https://csrc.nist.gov/projects/pec/zkproof).

Here is a proposed ecological example, not an implemented submission service: a provider wants to show that a calibrated instrument recorded dissolved oxygen above a specified threshold, in specified units, during an agreed interval, at a site within an approved region.

A proof design could keep the exact reading and location hidden while checking a defined relationship to a committed record. A commitment is a cryptographic way of fixing which hidden input is being used. The design would need to bind the relevant method, device identity, time window, units, threshold, and region to that same record.

The receiver would still need reasons to trust the instrument and the data collection. A valid computation over a fabricated reading remains a computation over fabricated input. Checking a threshold also does not establish why the condition occurred or whether an intervention caused it.

Before choosing a proof system, write the decision it would serve in a plain sentence. Then specify the evidence behind the inputs, what verification reveals, how fresh it must be, and how the same proof will be prevented from serving as a new event twice. Ask a cryptography specialist to examine the design and its assumptions.

## privacy lives in the surrounding system too

The right recipient list is only the start. Decide whether access includes downloading, further sharing, model processing, or publication of a derived conclusion. A reviewer may be permitted to inspect a recording while being unable to authorize its use in a public demonstration.

Check what other information a proposed summary exposes. A broad region combined with a precise time and a distinctive field photograph might still identify a site. Repeated queries or small groups can make apparently broad answers more revealing. A digest or hash used to identify a file should not be treated as an anonymity guarantee.

Set a review date for access and disclosure decisions. Agree on corrections, future access changes, retention, and any copy a recipient needs to keep for a legitimate audit. Be explicit about material that, once published, may be impossible to retrieve from other people's copies. Put sensitive raw material into a suitable controlled exchange rather than an irreversible public record.

These are practical design questions for ecological evidence, whether the contributor is a research laboratory, an acoustic-monitoring team, a community knowledge holder, or an API provider.

## make a contribution useful before sending the file

Start an initial conversation with four things:

1. The question the contribution could help answer.
2. The general kind of evidence and its coverage.
3. Who may authorize access and what restrictions apply.
4. A permitted public description or sample, if one is available.

For ensurance, private contribution arrangements and selective-proof options are proposed work to agree with contributors. The initial contact should contain a non-sensitive description. Discuss the appropriate exchange before sending exact species locations, restricted community knowledge, or confidential research.

That leaves room to choose a suitable response: a public summary, a controlled review, a more precise privacy design, or a request to keep the material with its holder. For the broader question of attaching a source to a living place, see [how to plug a sensor stack in](/guide/how-to-plug-a-sensor-stack-in?from=guide).

## frequently asked questions

### what is privacy-preserving ecological evidence?

It is evidence shared in a form that answers an intended question while limiting disclosure of sensitive material. The form might be a documented public summary, controlled access to a source, or a proof of a specific statement.

### does a zero-knowledge proof show that a sensor reading is true?

It verifies the mathematical relationship specified by the proof. Trust in the instrument, source identity, collection method, and physical measurement needs its own evidence. Ecological interpretation and causality require appropriate assessment too.

### can a sensitive wildlife location stay with its holder?

Yes. A collaboration can begin with a general description and agreed review terms. Decide what each task actually requires before choosing whether and how the precise record will be shared.

## discuss a private contribution

If your research or data could inform care for a place, tell us the question it addresses and what needs protection. Begin with a short non-sensitive description and a link you are allowed to share. We can then discuss an appropriate review and exchange.

[Discuss a private contribution](/contact?from=guide&topic=listening-across-species).

## sources

[U.S. Fish and Wildlife Service: Northern Spotted Owl](https://www.fws.gov/species/northern-spotted-owl-strix-occidentalis-caurina) — general species and habitat context.

[GBIF: Current Best Practices for Generalizing Sensitive Species Occurrence Data](https://docs.gbif.org/sensitive-species-best-practices/master/en/) — sensitivity assessment, documented generalization, and access decisions.

[NIST: Zero-Knowledge Proofs](https://csrc.nist.gov/projects/pec/zkproof) — the mathematical scope of zero-knowledge proofs.

## the series

Listening across species follows observations, interpretations and the care they could inform.

1. [what would it take to understand another species?](/guide/what-would-it-take-to-understand-another-species?from=guide)
2. [a whale calls back. what can we learn?](/guide/a-whale-calls-back-what-can-we-learn?from=guide)
3. [the plant has a signal. what does it mean?](/guide/the-plant-has-a-signal-what-does-it-mean?from=guide)
4. [one tree is a place to start](/guide/one-tree-is-a-place-to-start?from=guide)
5. [many ways of knowing, shared with care](/guide/many-ways-of-knowing-shared-with-care?from=guide)
6. [share the finding, protect the place](/guide/share-the-finding-protect-the-place?from=guide)
7. [from a signal to an act of care](/guide/from-a-signal-to-an-act-of-care?from=guide)
8. [bring one signal. help shape what happens next.](/guide/bring-one-signal-help-shape-what-happens-next?from=guide)