Home / Technology / AI Phishing: Scammers' New Trick No Typos Allowed
AI Phishing: Scammers' New Trick No Typos Allowed
18 Nov
Summary
- Generative AI now produces flawless phishing emails, removing old typo-based detection.
- AI enables personalized phishing by using public data for tailored lures.
- New AI defenses focus on request verification, not just message wording.
Generative AI has significantly advanced phishing tactics, rendering the long-standing advice of spotting typos and grammatical errors obsolete. Large language models can now produce fluent, error-free emails in various languages, making scams appear legitimate.
These AI tools also facilitate highly personalized phishing attacks. Scammers can leverage public profiles to craft individual lures, blurring the lines between mass and spear phishing campaigns. This personalization was previously labor-intensive.
Furthermore, AI enables scams in any language, removing a crucial detection layer for non-native speakers. It can mimic writing styles by analyzing existing email threads, making fraudulent messages indistinguishable from genuine correspondence.
Criminal-market chatbots like WormGPT and FraudGPT are advertised for creating phishing content. While mainstream chatbots have safety rules, scammers use jailbreak prompts or exploit loopholes. This means fraud-ready text generation is increasingly accessible.
AI also creates endless variations of phishing messages, slipping past spam filters that rely on repetition. It helps build convincing fake login pages, often indistinguishable from legitimate ones.
Phishing attacks now incorporate voice cloning and deepfake video calls. These synthetic media can back up fake email requests, making them seem more credible. For example, a $25 million fraud occurred after an employee joined a deepfake video call.
Effective defenses now focus on out-of-band verification, confirming requests through trusted channels separate from the suspicious message. Examining technical sender details, like the real domain, and using phishing-resistant sign-in methods such as passkeys are crucial. Email authentication protocols like SPF, DKIM, and DMARC also play a vital role in combating spoofing.