What This Tool Does
NoScamForMe is an AI-powered risk assessment tool. It analyzes text, emails, screenshots, URLs, and voice descriptions for patterns commonly associated with scams. It provides an informational risk assessment — not a guarantee of safety or danger.
Risk Assessment, Not Proof
A result of LOWER RISK does not prove a message is legitimate. It means the AI did not detect strong scam signals in the content provided. Sophisticated scams can evade detection. A result of SCAM does not guarantee the message is fraudulent — false positives occur.
False Positives and False Negatives
- False positives occur when a legitimate message is flagged as suspicious or a scam. This can happen with marketing emails, legitimate urgent notices, or messages using language similar to scam patterns.
- False negatives occur when a real scam is not detected. This is the more dangerous error. Novel scam types, well-crafted messages, or incomplete submissions can cause the AI to miss threats.
Content Quality Matters
Incomplete content reduces reliability. A single sentence from a longer conversation, a blurry screenshot, or a vague description may not contain enough information for a meaningful assessment. In these cases, the system may return UNABLE TO DETERMINE — this is an expected safety outcome, not a failure.
How the AI Models Work
Every check starts with Anthropic's Claude. For paid plans, any time there is uncertainty — and always before indicating lower risk — ChatGPT and Gemini provide independent assessments for comparison. Model agreement raises confidence but never proves a message is safe. Free plans use a single AI model.
The AI models compare your submission against a curated library of 3,001 documented scam patterns. The models are general-purpose AI systems, not purpose-built scam classifiers. They have broad knowledge but are not infallible.
LOWER RISK Does Not Mean Safe
"Lower risk" means no strong scam signals were detected in the content provided. It does not establish that the sender is legitimate, that the offer is real, or that acting on the message is safe. Always verify the sender through an independent, official channel before sending money or personal information.
UNABLE TO DETERMINE
This is an expected outcome when there is not enough reliable information to classify a message. It is a safety feature, not an error. Do not treat it as confirmation that a message is safe.
No Published Accuracy Rate
We do not publish an accuracy percentage because no formal validation study has been completed. We are developing a validation program using real-world scam samples and will publish results when available. Until then, treat every result as an informational data point, not a definitive answer.
The Bottom Line
NoScamForMe is a tool to help you think critically about suspicious messages. It is not a replacement for your own judgment, professional advice, or law enforcement. If something feels wrong, it probably is — regardless of what any AI says.
If you believe you have been scammed, contact the FTC at reportfraud.ftc.gov or call 1-877-382-4357.