Ethical Hacking News
Phishers have found a new use for invisible Unicode tag characters, a technique originally used to hide content from AI models, to evade detection in email phishing campaigns. As a result, defenders must adapt their strategies to counter this emerging threat and ensure that normalization and tokenization pipelines handle tag characters consistently.
Phishers are using ASCII smuggling, a technique originally associated with AI security risks, to evade detection and launch sophisticated email-based attacks. A massive phishing campaign was discovered, utilizing invisible Unicode tag characters to split financial-lure words in emails, evading keyword matching and content filters. The campaign's peculiar behavior, sending emails on weekdays and then going silent over the weekend, was a strategy to avoid detection by traditional security measures. The use of invisible Unicode tag characters highlights the versatility of ASCII smuggling and the need for defenders to be aware of its various applications. Defenders must verify that normalization and tokenization pipelines handle tag characters consistently to prevent ASCII smuggling attacks. Behavioral indicators, such as bulk volume and specific senders, can be used to detect and flag suspicious campaigns.
ASCII smuggling, a technique once primarily associated with AI security risks, has been repurposed by phishers to evade detection and launch sophisticated email-based attacks. In a recent revelation, Microsoft's security team discovered a massive phishing campaign utilizing invisible Unicode tag characters to split financial-lure words in emails, thereby evading keyword matching and other content filters. This new vector of attack highlights the ever-evolving nature of cybercrime and the need for defenders to remain vigilant and adapt their strategies to counter emerging threats.
The phishing campaign, which peaked at over 2.37 million messages in late February, remained elevated during weekdays over the next three months, only to decline significantly after May 15. The campaign's peculiar behavior – sending massive numbers of emails on weekdays and then going silent over the weekend – was noted by Microsoft's researchers. This pattern suggested that the attackers were attempting to avoid detection by traditional security measures.
The use of invisible Unicode tag characters in this campaign is reminiscent of ASCII smuggling, a technique used to hide content inside text that appears normal to humans. In this context, the phishers are using the invisible characters to split financial-lure words, making it difficult for content filters to detect the malicious intent. This is a stark contrast to the original use of ASCII smuggling, which involved hiding instructions for AI assistants within invisible Unicode characters.
Microsoft's researchers, Noam Kochavi and Sarah Wolstencroft, noted that the absence of instructions for AI assistants in the flagged messages suggested that the attackers were using the invisible tag characters for a different purpose. Instead, they were splitting financial-lure words to evade keyword matching and other content filters. This new use of ASCII smuggling highlights the versatility of this technique and the need for defenders to be aware of its various applications.
The researchers emphasized the importance of defenders verifying that normalization and tokenization pipelines handle tag characters consistently. They noted that any content that will be evaluated by keyword, signature, or regex logic should first have invisible and non-rendering Unicode code points stripped or folded, so that splicing them into a word no longer defeats the match. This control can also help reduce the threat of ASCII-smuggling against AI assistants that ingest email content.
In addition to verifying normalization and tokenization pipelines, defenders can also scan for behavioral indicators. The observed activity had a distinctive shape: bulk volume from churning, finance-themed disposable domains, on a strict weekday-on/weekday-off schedule. A sudden spike of tag-block characters concentrated on finance-themed senders, switching on and off weekly, is a high-confidence campaign indicator.
The revelation of this phishing campaign highlights the ever-evolving nature of cybercrime and the need for defenders to remain vigilant and adapt their strategies to counter emerging threats. As AI-era attack methods become better understood, threat actors may adapt them for use in more traditional threats such as phishing and spam. Defenders must view emerging threats through a cross-domain lens to stay ahead of the adversaries.
Related Information:
https://www.ethicalhackingnews.com/articles/Phishers-Hidden-Agenda-How-Invisible-Unicode-Tag-Characters-Became-a-New-Vector-for-Cybercrime-ehn.shtml
https://www.theregister.com/security/2026/09/04/ascii-smuggling-isnt-just-an-ai-security-risk/5294595
Published: Fri Sep 4 15:48:06 2026 by llama3.2 3B Q4_K_M
