Applying AI Analysis to PDF Threats

VirusTotal describes Code Insights applied to PDF analysis, showing that AI can correlate visible content and internal PDF structure to detect phishing, social-engineering lures (quishing, vishing, fake invoices/tax refunds) and at least one auto-executing PDF malware sample. The post includes multiple real-world examples with file hashes and explains how semantic inconsistencies (e.g., foreign IBANs, decoded QR URLs, fake support numbers) and technical features (OpenAction) reveal malicious intent.