Cybercriminals are increasingly using large language models to create highly convincing business email compromise (BEC) messages that closely resemble legitimate corporate communication, making them significantly more difficult to identify through traditional security methods. Unlike older phishing campaigns that often contained grammatical mistakes, suspicious links, or poorly written content, these AI generated emails are designed to appear professional and authentic, reducing the likelihood that recipients will recognize them as fraudulent. As attackers continue to adopt artificial intelligence to improve the quality of their social engineering campaigns, organizations face growing challenges in detecting malicious emails that contain no obvious indicators of compromise.
In response to this evolving threat landscape, Kaspersky has introduced new AI generated email detection technology within Kaspersky Security for Mail Server. The company said the technology is specifically designed to identify business email compromise attacks created with the help of large language models. Instead of relying solely on conventional indicators such as malicious links, suspicious attachments, or known phishing signatures, the detection engine uses artificial intelligence to analyze the content of emails and identify characteristics commonly associated with business email compromise attempts. According to Kaspersky, the system extracts BEC specific key phrases while also recognizing patterns that indicate machine generated text, allowing it to detect sophisticated fraudulent messages even when they contain no links or attachments. This approach is intended to strengthen email security against attacks that are increasingly difficult for both users and conventional filtering technologies to identify.
The company noted that the AI powered detection capability currently supports eight languages, enabling organizations operating across multiple regions to benefit from enhanced protection against AI generated email threats. Kaspersky also confirmed that support for additional languages is planned as the technology continues to evolve. The introduction of multilingual detection reflects the growing international nature of business email compromise campaigns, which frequently target organizations across different countries and industries. By analyzing linguistic patterns and contextual indicators instead of depending only on technical artifacts within an email, the technology aims to improve detection accuracy against increasingly sophisticated social engineering attempts. This development highlights how cybersecurity vendors are incorporating artificial intelligence into defensive technologies to counter attackers who are also leveraging AI to improve the effectiveness of their campaigns.
Business email compromise remains one of the most financially damaging forms of cybercrime because attackers typically impersonate trusted executives, business partners, or employees to convince recipients to transfer funds, disclose sensitive information, or perform unauthorized actions. As large language models continue to improve the quality and realism of malicious communications, organizations are expected to face more convincing email based threats that bypass traditional detection techniques. Kaspersky stated that its latest AI generated email detection capability is designed to address this challenge by identifying machine generated content alongside business email compromise indicators before such messages reach users. The company has published additional technical information explaining how the technology operates and how it enhances protection against modern AI assisted email attacks, providing organizations with greater insight into emerging risks associated with machine generated phishing and fraud campaigns.
Follow the SPIN IDG WhatsApp Channel for updates across the Smart Pakistan Insights Network covering all of Pakistan’s technology ecosystem.
