Attack Chain

1. An attacker positions themselves on the network path between a lettre client (built with boring-tls and using default-strict TLS settings) and its intended SMTP server.
2. The attacker obtains a chain-valid TLS certificate for a domain they control (e.g., via Let's Encrypt for attacker.example).
3. The vulnerable lettre client initiates an SMTP connection to its target mail exchanger (e.g., mail.example.com), configured to use TLS.
4. The attacker's intermediate SMTP server intercepts the connection and presents its certificate (e.g., for attacker.example) during the TLS handshake.
5. Due to the inverted boolean logic in lettre's boring-tls integration, the set_verify_hostname(false) call is made to the boring backend, despite the lettre client being configured for strict hostname verification (accept_invalid_hostnames = false).
6. The boring TLS backend, having been instructed to skip hostname verification, successfully completes the TLS handshake with the attacker's certificate, ignoring the mismatch with mail.example.com.
7. The lettre client proceeds with the SMTP transaction, sending sensitive data such as MAIL FROM, RCPT TO, message bodies, attachments, headers, and any SMTP AUTH credentials (e.g., PLAIN, LOGIN) to the attacker.
8. The attacker gains full read access to confidential communications and credentials, and can potentially manipulate messages before forwarding or dropping them.

Impact

This vulnerability severely impacts any lettre user deploying applications that utilize the boring-tls feature and rely on the default-strict hostname verification settings. Such users are unknowingly susceptible to Man-in-the-Middle attacks, allowing an on-path attacker to intercept all SMTP communications. This includes exposure of sensitive email content, attachments, and crucially, SMTP authentication credentials (PLAIN, LOGIN, CRAM-MD5, etc.). The confidentiality and integrity of email submissions are compromised, leading to potential data breaches, unauthorized access to email accounts, and impersonation. While Cloudflare independently identified and mitigated this downstream, the core vulnerability remains in affected lettre versions, requiring an upstream fix to protect all other deployments.

Recommendation

• Patch CVE-2026-46428 by updating the lettre library to version 0.11.22 or newer, which includes the fix.
• If immediate patching is not possible, avoid using the boring-tls feature with lettre in environments where SMTP traffic is sensitive and subject to MITM risks. Consider switching to native-tls or rustls backends if feasible, as they are unaffected by this specific bug.