Disclaimer

01 — What this service is.

EUDR Clear is decision-support tooling for EU Deforestation Regulation (Regulation (EU) 2023/1115 as amended by Regulation (EU) 2025/2650) preparation workflows. Every screening result, report, and dashboard carries this notice:

“Indicative screening for decision-support only — not a compliance determination, certification, or legal advice. Due-diligence obligations remain with the operator (EUDR Art. 10).”

02 — Verdict vocabulary.

Our outputs use a fixed, mechanical vocabulary. Each verdict chip describes only what the pinned datasets show for your plot:

• NO SIGNAL DETECTED — the screened datasets, at the pinned versions, show no flagged signal for this plot. This is not proof of the absence of deforestation, and not proof that the plot is production-free under the EUDR.
• REVIEW RECOMMENDED — at least one dataset shows a signal that warrants human review (for example, tree-cover loss pixels near the boundary, or alerts in the buffer zone).
• ELEVATED SIGNAL — multiple datasets or a stronger signal class coincide for this plot. This is a prioritisation cue for review, not a conclusion about illegality.
• DEMO — output produced by the deterministic demo mode; watermarked and not for regulatory use.

We never mark a plot "compliant", "approved", or verified. There is no verdict chip for that, by design.

03 — Dataset limitations and error rates.

Screening is only as good as the upstream datasets, and every dataset has known error characteristics:

• GLAD/RADD forest alerts are near-real-time detection models. Published validation studies report both false positives (alerts without true disturbance; e.g. from clouds, shadows, fire, or seasonal wetland changes) and false negatives (disturbance that generates no alert; especially under persistent cloud cover, in dry forests, and outside the humid tropics). Current published user's-accuracy figures: [XX]% humid tropics · XX% dry tropics — cite the UMD GLAD lab's latest validation page and date].
• Hansen Global Forest Change has 30 m resolution; small-scale disturbance, degradation, and events late in the annual update cycle can be missed, and year assignments carry uncertainty.
• FAO Open Foris Whisp indicators inherit the accuracy limits of their input layers.
• Copernicus Sentinel-2 observations are limited by cloud cover and revisit cadence; a "no signal" state can reflect missing observations, not calm ground truth.

Each result lists the exact dataset versions ("vintages") used, so any figure can be traced and re-run.

04 — Not certification, not legal advice.

EUDR Clear does not certify, attest to, or approve anything, and does not provide legal advice. We do not issue audit opinions. If a counterparty, auditor, or authority asks whether your product is compliant, no output of this service answers that question.

05 — Operator obligations.

Collecting information, exercising due diligence, submitting due diligence statements, and keeping records remain the operator's obligations under the EUDR (Arts. 3, 4, 10, and 12). Our evidence packs are inputs to that work, not a substitute for it.

06 — Independence.

EUDR Clear is an independent commercial product. It is not affiliated with, sponsored by, or endorsed by the European Commission, the European Union, the FAO, or any dataset provider. Where this site simplifies the regulation, the regulation prevails.

07 — Demo mode.

When the live Whisp API is not configured, the engine runs a documented deterministic demo mode so the workflow can be evaluated end-to-end. Demo outputs are always watermarked DEMO and are not fit for any regulatory use.