Automated voicemail transcripts have become part of the daily rhythm of enterprise communication. They arrive with familiar subject lines, system-generated formatting, and little human context, exactly the qualities that make them easy to trust and easy to overlook. Attackers are now exploiting that familiarity, turning routine call-transcription notifications into a vehicle for credential phishing.

Check Point researchers identified a large-scale phishing campaign that exploits this shift in enterprise behavior. The emails impersonate automated voicemail-transcript notifications and deliver malicious Scalable Vector Graphics (SVG) attachments that redirect users to credential-harvesting pages, converting a familiar collaboration workflow into a potential path for account takeover.

A Phishing Campaign Built to Look Routine

Between August 17 and August 31, Check Point identified more than 58,000 emails tied to the campaign. The operation targeted over 7,800 organizations, leveraging more than 38,400 spoofed sender addresses across over 9,300 spoofed domains.

The subject lines follow a simple formula. Each begins with “Automated transcript,” followed by a partially redacted phone number and a random tracking string. The effect is deliberately understated: a notification that appears to have been generated by a trusted workplace system.

Inside the Attack Chain

The campaign leans heavily on trust signals that are difficult for users to evaluate quickly. Each email spoofs the recipient’s own domain as the sender, giving the message the appearance of an internal system alert before the attachment is even opened.

The attachment is designed to appear consistent with a call-recording workflow, using names such as “▷ ——— 001min 09sec_….svg.” In reality, the file is an SVG containing embedded script. When opened, the script redirects the user to a phishing page. Because the user’s email address is already encoded in the URL, the fake login form can auto-fill it, creating a more personalized and credible credential-harvesting experience.

Notable Characteristics

  • Sender addresses are spoofed to match the recipient’s own domain, making the message appear internal.
  • Subject lines mimic notifications produced by real voicemail and call-transcription agents.
  • The payload is delivered as an SVG rather than a traditional executable or macro-enabled document.
  • The redirect executes client-side, leaving no visible link for the recipient to inspect before opening the attachment.
  • The landing page is pre-filled with the victim’s email address, lowering friction and making the fake login prompt appear more credible.

Where Check Point Email Security Stops the Attack

This campaign illustrates the limitations of defenses that rely primarily on reputation, known malicious links, or obvious attachment types. The sender is spoofed, the lure is familiar, the payload is an SVG rather than a traditional executable, and the malicious destination is reached only after the file is opened.

Check Point Email Security is designed to stop these threats before they reach the inbox. Powered by ThreatCloud AI and more than 60 AI engines, it analyzes sender behavior, message context, attachment characteristics, URL intent, and phishing signals across the attack chain. This multi-layered analysis is critical when attackers use randomized infrastructure, client-side redirects, and file formats that appear harmless to conventional filters.

In campaigns like this, Check Point Email Security evaluates the spoofed sender domain, automated-transcript lure, suspicious SVG attachment, embedded redirect behavior, and credential-harvesting landing page. By correlating these signals, it can help prevent message delivery, quarantine malicious attachments, and block access to phishing sites if the attack progresses through a browser session.

The Bigger Risk: Automated Trust

The campaign is not only a phishing case study. It reflects how quickly attackers adapt to enterprise workflows as automation becomes more common. AI agents are increasingly being asked to triage inboxes, summarize messages, process attachments, and initiate follow-up actions on behalf of users. As that responsibility expands, the same social-engineering cues designed to influence employees can also introduce risk into automated workflows if appropriate controls are not in place.

For example, an agent or automated workflow that is permitted to open attachments or follow links may process a file named, “call recording” without the same contextual judgment expected from a security-aware user. This scenerio highlights the need to apply enterprise controls, access limits, and monitoring to automated systems that handle inbound content.

How Organizations Can Reduce Risk

  • Treat “automated” notifications as a signal to verify, not a reason to relax scrutiny, especially when the sender appears to match your own domain.
  • Before deploying an AI agent to open attachments, follow links, or act on inbound content, define which file types and domains it can access without human confirmation.
  • Inspect SVG attachments as active content, not simply as images.
  • Log and review agent actions the same way you would monitor privileged user activity, so misuse or compromise does not go unnoticed.
  • Encourage users to report suspicious automated notification emails so security teams can identify related infrastructure and block follow-on attempts.

You may also like