DarkwaterDARKWATER← Back to home

Legal

Darkwater AI Guide

Last updated: July 1, 2026

Darkwater uses generative AI to help you screen vessels and counterparties, surface risk signals, and turn raw maritime data into readable analysis faster. This guide explains what our AI features do, where their limits are, and what we expect from you when you rely on them. It supplements our Terms of Use and Privacy Policy.

1. Overview and purpose

Some Darkwater features are powered by large language models and related AI systems, run through established providers. These systems operate over Darkwater’s own data and domain knowledge, licensed and open-source intelligence (OSINT), and — where you provide them — your own inputs, to generate summaries, explanations, and suggested next steps. AI is one input into a workflow that is designed around human review.

2. How Darkwater uses AI

  • Risk summaries — plain-language explanations of why a vessel or counterparty was flagged, drawing on ownership, behaviour, and sanctions signals.
  • Adverse-media & screening support — surfacing and summarising relevant negative news and source material.
  • Drafting assistance — helping draft briefs, notes, and communications you can review and edit.
  • Guided analysis — multi-step reasoning that helps assemble ownership chains and connect related entities.

3. Accuracy and limitations

AI outputs can vary in quality and may contain inaccuracies, omissions, or outdated information, and can occasionally present incorrect statements with apparent confidence (“hallucinations”). Outputs should be treated as general reference material to be verified — not as definitive facts, legal or compliance advice, or a substitute for professional judgement.

4. Human oversight and your responsibilities

Darkwater is decision-support software: a human must remain in the loop. Before you act on any AI-assisted output, you are responsible for:

  • independently verifying material facts against primary sources;
  • applying your own professional and regulatory judgement;
  • ensuring your use is lawful, ethical, and appropriate to the decision at stake;
  • owning the final go/no-go, compliance, credit, or commercial decision.

5. Acceptable use

When using Darkwater AI features, you must not:

  • submit content that infringes intellectual-property or privacy rights;
  • upload malicious content or attempt to bypass safety controls or guardrails;
  • enter sensitive personal data that is not necessary for your legitimate use;
  • use outputs to unlawfully discriminate, harass, or cause harm;
  • present AI-generated output as independently verified where it is not.

6. How your data is handled

We do not use your inputs or outputs to train third-party foundation models, and our AI providers are contractually restricted from using your inputs or outputs to train their models. Your data is processed to deliver the feature you requested and is handled in line with our Privacy Policy.

7. Language and liability

AI output — including any non-English translations — may contain errors. To the extent permitted by law, Darkwater is not liable for the accuracy of AI output or for decisions made in reliance on it. See our Terms of Use for full details.

8. Contact us

Questions about how Darkwater uses AI? Contact us at darkwaterlabs@proton.me.