AI chatbots are being manipulated to redirect users to scams. A detection API could analyze conversation patterns, URLs, and context to flag suspicious activity.
Offer it as a simple REST endpoint that checks text exchanges. Charge per API call with volume discounts.
Start with basic pattern matching and URL analysis. Later add ML models trained on scam examples.
The risk is false positives blocking legitimate conversations. Start with low-confidence warnings rather than hard blocks.