Google Notification alert link from non-Google sender

 1name: "Google Notification alert link from non-Google sender"
 2description: "This rule detects messages that leverage a link to notifications.google.com not from google and from an untrusted sender. Commonly abused in salesforce phishing campaigns. "
 3type: "rule"
 4references:
 5  - "https://www.reddit.com/r/Scams/comments/15oilcr/meta_business_suite_messenger_phishing_scam_can/"
 6severity: "medium"
 7source: |
 8  type.inbound
 9  // ignore messages from google[.]com unlesss they fail DMARC authentication
10  and (
11    (
12      sender.email.domain.root_domain in ("google.com", "youtube.com")
13      and not headers.auth_summary.dmarc.pass
14    )
15    or sender.email.domain.root_domain not in ("google.com", "youtube.com")
16  )
17  and any(body.links,
18          .href_url.domain.domain == "notifications.google.com"
19          and strings.starts_with(.href_url.path, "/g/p/")
20  )
21  and any($suspicious_subjects, strings.icontains(subject.subject, .))
22  and (
23    (
24      profile.by_sender().prevalence in ("new", "outlier")
25      and not profile.by_sender().solicited
26    )
27    or (
28      profile.by_sender().any_messages_malicious_or_spam
29      and not profile.by_sender().any_false_positives
30    )
31    // we've observed salesforce abuse
32    or sender.email.domain.root_domain == "salesforce.com"
33    or headers.return_path.domain.root_domain == "salesforce.com"
34  )
35  and not profile.by_sender().any_false_positives  
36attack_types:
37  - "Credential Phishing"
38tactics_and_techniques:
39  - "Social engineering"
40detection_methods:
41  - "Content analysis"
42  - "Header analysis"
43  - "Sender analysis"
44  - "URL analysis"
45id: "a1c1acfd-9b3b-58a7-81f1-b2c631b02985"