AI Transparency

How FraudSentry uses AI without hiding the limits

FraudSentry keeps the product-owned assessment layer separate from AI-generated sections so people can understand what they are reading, what still needs checking, and where official support still matters.

FraudSentry assessment

This is the platform-owned assessment layer based on user-provided information, structured warning signs, rules, patterns, and product logic. It explains why FraudSentry became concerned, but it is still decision-support only and not proof.

LLM agent review

This section explains the result in plain language to help people understand the situation more quickly. It is clearly labelled as an LLM agent review, may be inaccurate, and does not turn the report into anything official, certified, or final.

AI agent guidance

This section suggests practical next steps to help you act more safely. It is AI agent guidance, not legal advice, financial advice, emergency services, formal investigation, official evidence, or proof for any institution.

What FraudSentry helps with

Helping people slow down and verify more safely

FraudSentry helps people assess suspicious contact, understand warning signs, and find safer next steps. It is designed to support judgment, not replace it.

The product is broader than a simple message checker. It is meant to help across calls, texts, emails, suspicious links, payment pressure, and country-aware support needs.

Its strongest role is helping people pause, compare the contact with known warning signs, and move toward official or trusted verification routes instead of following the suspicious contact's instructions.

What FraudSentry does not do

It does not remove the need for official help

FraudSentry does not guarantee that a message, caller, or website is legitimate or fraudulent.
FraudSentry does not replace your bank, police, consumer authorities, regulator, or emergency services.
FraudSentry should not be used as the only reason to send money, share information, or delay urgent official action.
FraudSentry reports and AI sections should not be treated as evidence, proof, a certified decision, or a formal investigative output for any institution.

Plain-English limit

AI support does not make anything official

The FraudSentry assessment, LLM agent review, and AI agent guidance can help explain risk and next steps, but none of them are official proof, a police report, a bank decision, legal advice, financial advice, emergency support, court-ready evidence, or proof for any institution. Verify important actions independently through official or trusted channels.

Trust principles

What to keep in mind when reading a report

AI-generated sections may be inaccurate.
Reports are guidance, not proof, official evidence, or certified findings.
Independent verification still matters.
Official services still matter.
Do not use suspicious links, phone numbers, QR codes, or instructions to verify.

Independent verification

When official services matter most

If money has already been sent, accounts may be compromised, codes were shared, or remote access software was installed, contact the relevant official organisation directly using details you already trust.

FraudSentry can help organise the situation and point you toward safer next steps, but it should not delay urgent contact with your bank, police, or official support service.