A European security researcher recently discovered that current implementations of the RFC-1342 standard, which was created in 1992 to encode non-ASCII characters inside email headers, may have security vulnerabilities that enable fraudsters to commit spoofing and code injection attacks.
These flaws allow spoofers to circumvent Domain-Based Message Authentication, Reporting and Conformance (DMARC), an antispoofing authentication protocol introduced in 2015, as well as spam filters, the researcher, Sabri Haddouche, wrote in a blog post.
Poking Holes in RFC-1342
RFC-1342 converts non-ASCII characters in mail headers to standard ASCII characters, which is required by most email servers. But Haddouche discovered that some email services would take an RFC-1342-encoded string and decode it without checking for objectionable code that could be embedded in the string.
Due to the way email clients display sender addresses, DMARC would then evaluate the DomainKeys Identified Mail (DKIM) signature of the original domain as correct. This makes phishing emails and their improbable addresses harder for defenders to spot.
Additionally, Haddouche noted that if the decoded email string contained a null byte or multiple email addresses, things went haywire. The client would read only the email address before the null byte or the first valid email found and ignore the rest of the encoded string. An attacker could fool an email client by generating a non-ASCII string to serve as a sender address with these properties.
Addressing Security Vulnerabilities in Email Clients
Haddouche compiled a list of vulnerable email clients, which includes clients for all major consumer operating systems. Although he contacted all 33 affected vendors, Bleeping Computer reported than only eight have released patches to fix the problem. Haddouche noted that Mozilla and Opera said they won’t address the bug because they consider it to be a server-side issue.
IT professionals must consider the security of the email systems they use in the enterprise. To address this specific threat, security teams may need to contact affected vendors to reach a resolution.