Service abuse: Zohodesk reply-to mismatch with job scam indicators

calendar Jul 22, 2026  ·
Share on: linkedin copy

Detects inbound messages sent from Zohodesk infrastructure where the reply-to address points to a domain outside of Zohodesk, combined with natural language signals indicating job scam content. This technique abuses legitimate Zohodesk services to add credibility while redirecting responses to an external actor-controlled address.

Sublime rule (View on GitHub)

 1name: "Service abuse: Zohodesk reply-to mismatch with job scam indicators"
 2description: "Detects inbound messages sent from Zohodesk infrastructure where the reply-to address points to a domain outside of Zohodesk, combined with natural language signals indicating job scam content. This technique abuses legitimate Zohodesk services to add credibility while redirecting responses to an external actor-controlled address."
 3type: "rule"
 4severity: "medium"
 5source: |
 6  type.inbound
 7  and sender.email.domain.root_domain == "zohodesk.com"
 8  and any(headers.reply_to, .email.domain.root_domain != "zohodesk.com")
 9  and (
10    any(ml.nlu_classifier(body.current_thread.text).intents,
11        .name == "job_scam" and .confidence != "low"
12    )
13    // nlu fallback where we don't get job scam
14    or strings.icontains(body.current_thread.text, "talent acquisition")
15  )  
16attack_types:
17  - "BEC/Fraud"
18  - "Spam"
19tactics_and_techniques:
20  - "Social engineering"
21  - "Out of band pivot"
22  - "Spoofing"
23detection_methods:
24  - "Header analysis"
25  - "Sender analysis"
26  - "Natural Language Understanding"
27id: "af2ab304-7f1e-5284-936d-f04ae91370f7"
to-top