Spam Detection

Screens incoming email for phishing and forwards suspicious messages to security analysts.

Business impact

• phishing detection precision — Nestlé reported 60 percent precision for the phishing detector.

Data requirements

• incoming email (text) — The detector screens email messages for phishing indicators.
• new attack examples (text) — The model retrains on newly observed attacks.

AI methods and techniques

• machine learning — Used to screen incoming email for phishing.
• deep learning — The article describes an AI-powered phishing detector built with a transformer model.
• retraining on new attacks — The system is updated with newer attack patterns so detection stays current.

AI models and model families

transformer model

Industries

Real-world evidence

2 documented case studies on record.

Companies using this: Nestlé, Bharti Airtel.