AtData Introduces Gibberish Detection to Identify Bot-Generated Emails

AtData has launched Gibberish Detection, a new machine learning-driven feature in its fraud prevention suite, announced in a press release. The system identifies and intercepts email addresses that appear synthetic, random, or automated at the point of capture.

The model analyzes the structure of an email address to detect signs of randomness or automation, using indicators such as pattern anomalies and likely bot behavior. It provides a real-time, confidence-weighted signal that can block or score suspicious inputs instantly.

AtData reports that about 5% of email inputs across its network appear to be gibberish or randomly generated, and in some deployments, that rate can reach 10%. The company says the new feature helps fraud and risk teams make faster, more accurate decisions while reducing manual review costs by filtering low-value registrations early in the process.

Gibberish Detection is now available through AtData's Fraud API endpoint.