Microsoft 365 Service Health Impersonation and Account Abuse
An intruder has gained initial access by sending phishing emails that impersonate Microsoft 365 outage notifications, inducing users to click tracking links that lead to credential theft and subsequent unauthorized cloud API activity.
Based on research by Microsoft 2026-11-17 11 steps · 3 queries T1036 T1041 T1071 T1190 T1204.001 T1566 T1566.002
Brief
Why now
Microsoft Security recently detailed a "High volume phishing campaign uses Microsoft 365 outage lures to target multiple industry sectors" (https://www.microsoft.com/en-us/security/blog/2026/11/16/high-volume-phishing-campaign-uses-microsoft-365-outage-lures-target-multiple-industry-sectors-2/). The adversary sends realistic notifications about service outages or security updates to induce users to click tracking links. These links lead to credential theft and subsequent business email compromise (BEC) activity.
How the hunt flows
The first phase uses a cheap lead query against Microsoft 365 mail telemetry. It searches for specific sender domains and subject lines associated with the service-health lures. This initial step gates the rest of the hunt to ensure we only spend resources on expensive telemetry when relevant lures exist in the environment.
If the lead query returns matches, the hunt fans out into a parallel search across two distinct surfaces. The first branch examines DNS activity to find hosts that resolved the campaign's known tracking and redirection domains. This provides evidence that a user actually clicked the link within the phishing email.
At the same time, the second branch searches cloud API logs for rare impact indicators. It looks for the creation of new inbox rules or OAuth application consents, filtering for operations that occur across only a small number of users. This step identifies the actual impact of the credential theft.
In the final phase, an analyst or automated agent correlates these three points: the delivery of the email, the interaction with the link, and the subsequent API abuse. This sequence confirms a successful account takeover rather than just a blocked phishing attempt.
What this hunt cannot see
This hunt relies on the Microsoft 365 Unified Audit Log. If logging is not centralized or certain mailbox types are excluded, the lead query will not see the incoming lures. Furthermore, the hunt uses a fixed lookback window. If the adversary waits longer than 14 days between the link click and the API abuse, the correlation may fail depending on your local log retention settings.
Steps
-
Lead: Outage-themed phishing lures
Query · scopingIdentify suspicious emails using the campaign's specific sender domains and subject line patterns.
reads microsoft365_mail_messagesqlSELECT sender, "from", subject, sent_date_time, internet_message_id, tenant_id FROM microsoft365_mail_message WHERE (instr(',' || '{{sender_domains}}' || ',', ',' || LOWER(SUBSTR(sender, INSTR(sender, '@') + 1)) || ',') > 0 OR subject LIKE 'Outage -%' OR subject LIKE 'Strengthen - Security%' OR subject LIKE 'Restore — 4 messages%') AND sent_date_time >= datetime('now', '-{{lookback_days}} days')What a hit looks like. Rows identify specific phishing emails delivered to users. Silence suggests the campaign has not targeted the tenant in the current window.
-
Verify phishing campaign relevance
Agent triageConfirm if the detected emails match the high-volume outage campaign profile based on the reported lures.
-
Gate on phishing evidence
DecisionRoute to expensive telemetry queries only if phishing emails are confirmed.
-
Scoped DNS interaction with tracking domains
Query · enrichmentVerify if scoped hosts resolved the campaign's tracking or redirection domains, indicating a link click.
reads hb_dns_activitysqlSELECT device_hostname, query_hostname, answers, time FROM hb_dns_activity WHERE (('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0)) AND instr(',' || '{{tracking_domains}}' || ',', ',' || LOWER(query_hostname) || ',') > 0 AND time >= datetime('now', '-{{lookback_days}} days')What a hit looks like. Resolution of tracking domains by a host suggests a successful link click from a phishing email.
-
Rare cloud API inbox rule manipulation
Query · detection candidateDetect account impact by finding users performing rare inbox rule operations or OAuth consents typical of BEC.
reads hb_cloud_api_activitysqlSELECT api_operation, actor_user_name, COUNT(DISTINCT actor_user_name) AS user_count, MIN(time) AS first_seen FROM hb_cloud_api_activity WHERE provider = 'm365' AND (api_operation IN ('New-InboxRule', 'Set-InboxRule', 'Consent to application')) AND time >= datetime('now', '-{{lookback_days}} days') GROUP BY api_operation, actor_user_name HAVING user_count <= 3What a hit looks like. Specific users creating inbox rules when such behavior is rare across the tenant. Silence suggests no automated rule manipulation has occurred.
-
Correlate email, interaction, and abuse
Agent triageAn agent weighs the email receipt, DNS interaction, and rare API activity to confirm successful compromise.
-
Route on compromise verdict
DecisionInitiate remediation if an active compromise is confirmed.
-
Isolate compromised account
Response actionContain the impact by revoking sessions and rotating credentials.
-
Forensic analyst review
Analyst taskVerify the automated correlation and determine the extent of data access.
-
Close out hunt
Analyst taskFinalize documentation and record any identified blind spots.
Coverage
Scenario coverage
| Stage | Covered | How, or why not |
|---|---|---|
| Outage-Themed Phishing Delivery T1566.002 · T1566 · T1036 |
Yes | phishing-emails-lead |
| Malicious Link Redirection T1204.001 · T1071 · T1190 |
Yes | dns-tracking-interaction |
| Assessed Account Abuse T1041 |
Yes | rare-cloud-api-impact |
Blind spots
- Needs Complete M365 Unified Audit Log ingestion. If certain mailbox types or external tenants are not integrated, the lead query will return zero rows even if phishing was successful. It would answer whether all delivered phishing emails were visible to the lead query. Remediation: Ensure all M365 tenants are enrolled in centralized audit logging.
- Needs hb_cloud_api_activity retention > 14 days. A compromise that happened 20 days ago would be missed if logs roll over quickly. It would answer whether the impact occurred outside the current lookback window. Remediation: Increase cloud API log retention to at least 30 days.
Parameters & data
Parameters
| Parameter | Type | Default | What it is |
|---|---|---|---|
lookback_days | number | 14 | Days of history to examine. |
scope_hosts | list[host] | — | Optional list of hostnames to narrow the expensive DNS query; leave empty for fleet-wide. |
sender_domains | list[domain] | moments-notice.com, planetmaison.com, newbooker.com, em4493.moments-notice.com, em.newbooker.com | Observed sender domains from the campaign report. |
tracking_domains | list[domain] | u113727626.ct.sendgrid.net, u118254266.ct.sendgrid.net, u118250677.ct.sendgrid.net, url.newbooker.com, url7489.newbooker.com, reallifeaccounting.com, outagefixed.sbs, outagerestored.sbs | Infrastructure used for click-tracking and redirection. |
Telemetry
| Source | Category | Telemetry |
|---|---|---|
| Endpoint telemetry (hb_ surfaces) | endpoint | endpoint |
| Identity / sign-in telemetry | identity | identity |
Source
---
analysis: A standard rule would alert on the phishing email, but this hunt correlates
the delivery with specific host interaction (DNS) and specific account impact (rare
API calls) in one flow, providing the end-to-end evidence an analyst needs to confirm
an intrusion.
blind_spots:
- id: mail-log-visibility
owner: Messaging Team
question: whether all delivered phishing emails were visible to the lead query
remediation: Ensure all M365 tenants are enrolled in centralized audit logging.
requires: Complete M365 Unified Audit Log ingestion
risk: If certain mailbox types or external tenants are not integrated, the lead
query will return zero rows even if phishing was successful.
stage: initial-access-service-outage-phishing
- id: no-api-retention
owner: Platform Engineering
question: whether the impact occurred outside the current lookback window
remediation: Increase cloud API log retention to at least 30 days.
requires: hb_cloud_api_activity retention > 14 days
risk: A compromise that happened 20 days ago would be missed if logs roll over quickly.
stage: impact-account-compromise-activity
coverage:
- stage: initial-access-service-outage-phishing
status: covered
steps:
- phishing-emails-lead
- stage: execution-tracking-link-interaction
status: covered
steps:
- dns-tracking-interaction
- stage: impact-account-compromise-activity
status: covered
steps:
- rare-cloud-api-impact
guardrails:
claims: no_unsupported
evidence: citation_required
missing_data: not_benign
telemetry: untrusted
hunt:
applicability: campaign-specific
handoff: promote-to-detection
justification: Credential phishing impersonating platform outages has a high success
rate. Confirming a negative result across the estate ensures the perimeter blocks
were effective, while a positive result enables containment before mailbox data
is exfiltrated.
methodology: model-assisted
trigger: intel-report
hypothesis: An intruder has gained initial access by sending phishing emails that
impersonate Microsoft 365 outage notifications, inducing users to click tracking
links that lead to credential theft and subsequent unauthorized cloud API activity.
labels:
- hunt
- attack.t1566.002
- attack.t1566
- attack.t1036
- attack.t1204.001
- attack.t1071
- attack.t1190
- attack.t1041
- execution
- impact
- initial access
- microsoft-365
name: Microsoft 365 Service Health Impersonation and Account Abuse
parameters:
lookback_days:
default: '14'
description: Days of history to examine.
from:
kind: manual
observed: '2026-11-17'
ref: huntbase-standard
type: number
scope_hosts:
default: []
description: Optional list of hostnames to narrow the expensive DNS query; leave
empty for fleet-wide.
from:
kind: manual
ref: analyst-defined
type: list[host]
sender_domains:
default:
- moments-notice.com
- planetmaison.com
- newbooker.com
- em4493.moments-notice.com
- em.newbooker.com
description: Observed sender domains from the campaign report.
from:
kind: article
observed: '2026-11-13'
ref: ms-blog-2026-11
type: list[domain]
tracking_domains:
default:
- u113727626.ct.sendgrid.net
- u118254266.ct.sendgrid.net
- u118250677.ct.sendgrid.net
- url.newbooker.com
- url7489.newbooker.com
- reallifeaccounting.com
- outagefixed.sbs
- outagerestored.sbs
description: Infrastructure used for click-tracking and redirection.
from:
kind: article
observed: '2026-11-13'
ref: ms-blog-2026-11
type: list[domain]
provenance:
authors:
- name: Huntbase hunt generation
org: huntbase.io
generated:
by: huntbase-hunt-generation
from: https://www.microsoft.com/en-us/security/blog/2026/11/16/high-volume-phishing-campaign-uses-microsoft-365-outage-lures-target-multiple-industry-sectors-2/
gates:
- dry-run
- lint
- critic
model: hb_google/gemini-3-flash-preview
rationale: The lead query is a cheap search on mail telemetry for specific lures.
Because the campaign is high-volume and uses specific domains, we start broad across
all mailboxes. Scoping is only applied at the DNS level if an analyst needs to re-run
for specific hosts identified in the lead.
references:
- name: "High volume phishing campaign uses Microsoft 365 outage lures \u2014 Microsoft\
\ Security Blog"
url: https://www.microsoft.com/en-us/security/blog/2026/11/16/high-volume-phishing-campaign-uses-microsoft-365-outage-lures-target-multiple-industry-sectors-2/
related:
- hunt: m365-inbox-rule-manipulation-baseline
reason: This hunt focuses on phishing-led rule changes; a sibling hunt baselines
all rule changes regardless of vector.
relation: sibling
scenario:
stages:
- name: Outage-Themed Phishing Delivery
observables:
- 'Sender domains: moments-notice.com, planetmaison.com, newbooker.com, em4493.moments-notice.com,
em.newbooker.com'
- "Subject patterns: 'Outage \u2013 ', 'Strengthen \u2013 Security ', 'Restore\
\ \u2014 4 messages '"
- Display names ending in 'Outage Notification' or 'Service Health'
- Lure text referencing Exchange Online, Outlook, Teams, SharePoint, and Copilot
- Urgency claims of 24-hour message deletion
slug: initial-access-service-outage-phishing
tactic: initial-access
techniques:
- T1566.002
- T1566
- T1036
- name: Malicious Link Redirection
observables:
- 'Click-tracking domains: u113727626.ct.sendgrid.net, u118254266.ct.sendgrid.net,
u118250677.ct.sendgrid.net'
- 'Tracking subdomains: url.newbooker.com, url7489.newbooker.com'
- 'Destination domains: reallifeaccounting.com, outagefixed.sbs, outagerestored.sbs'
- 'URL paths: /themelist0.html, /topics_pxyk/themelist0.html'
- 'Attacker IP: 162.120.17.69'
slug: execution-tracking-link-interaction
tactic: execution
techniques:
- T1204.001
- T1071
- T1190
- name: Assessed Account Abuse
observables:
- Sign-ins from 162.120.17.69
- Creation of new inbox or forwarding rules
- OAuth application consent operations
- Anomalous access to SharePoint or Teams data following login
slug: impact-account-compromise-activity
tactic: impact
techniques:
- T1041
summary: A high-volume phishing campaign leveraged Microsoft 365 outage-themed lures
to trick users into clicking recipient-specific tracking links. These links redirected
victims to a common destination infrastructure for assessed credential phishing,
often referencing Exchange Online disruptions and requiring immediate action to
restore 'held' messages.
severity: high
targets:
analyst:
name: Tier-2 analyst
role: analyst
endpoint:
category: endpoint
name: Endpoint telemetry (hb_ surfaces)
telemetry:
- endpoint
hunter:
agent: true
name: Hunt agent
identity:
category: identity
name: Identity / sign-in telemetry
telemetry:
- identity
microsoft-365:
category: siem
huntbase:
product: microsoft-365
name: microsoft 365
tlp: clear
type: investigation
---
# Microsoft 365 Service Health Impersonation and Account Abuse
This hunt targets a coordinated phishing campaign using service-health lures. It employs a gated flow where a cheap lead query on mail telemetry determines if campaign-specific lures exist. If confirmed, the hunt fans out to verify user interaction via DNS tracking infrastructure and identifies post-compromise impact through rare cloud API operations, such as the creation of new inbox rules or OAuth application consents. An agent correlates the delivery, interaction, and impact to confirm an active intrusion.
## phishing-emails-lead
<!-- Lead: Outage-themed phishing lures -->
Identify suspicious emails using the campaign's specific sender domains and subject line patterns.
```sqlite target=microsoft-365 role=scoping params=(sender_domains=sender_domains, lookback_days=lookback_days)
~~~yaml
expected: Rows identify specific phishing emails delivered to users. Silence suggests
the campaign has not targeted the tenant in the current window.
reads:
- sender
- from
- subject
- sent_date_time
- internet_message_id
- tenant_id
silence: not_evidence_of_absence
source: microsoft365_mail_message
verified: dry-run
verified_at: '2026-11-17'
~~~
SELECT sender, "from", subject, sent_date_time, internet_message_id, tenant_id FROM microsoft365_mail_message WHERE (instr(',' || '{{sender_domains}}' || ',', ',' || LOWER(SUBSTR(sender, INSTR(sender, '@') + 1)) || ',') > 0 OR subject LIKE 'Outage -%' OR subject LIKE 'Strengthen - Security%' OR subject LIKE 'Restore — 4 messages%') AND sent_date_time >= datetime('now', '-{{lookback_days}} days')
```
## verify-lead
<!-- Verify phishing campaign relevance -->
```agent target=hunter
cite: required
context:
- phishing-emails-lead
max_iterations: 3
objective: Determine if the identified mail records represent the Microsoft 365 outage
phishing campaign.
success_criteria: A per-row verdict of malicious or benign.
tools:
- endpoint
- microsoft-365
```
## is-lead-legit
<!-- Gate on phishing evidence -->
if~: "the verify-lead verdict is malicious for at least one email record" (confidence: high, judge=hunter)
then: → fan-out-expensive
indeterminate: → analyst-review
unavailable: → analyst-review (blind_spot: mail-log-visibility)
else: → close-out
## fan-out-expensive
<!-- Expensive correlation fan-out -->
parallel:
- → dns-tracking-interaction
- → rare-cloud-api-impact
join: → triage-correlation
## dns-tracking-interaction
<!-- Scoped DNS interaction with tracking domains -->
Verify if scoped hosts resolved the campaign's tracking or redirection domains, indicating a link click.
```sqlite target=endpoint role=enrichment params=(scope_hosts=scope_hosts, tracking_domains=tracking_domains, lookback_days=lookback_days)
~~~yaml
expected: Resolution of tracking domains by a host suggests a successful link click
from a phishing email.
reads:
- device_hostname
- query_hostname
- answers
- time
silence: not_evidence_of_absence
source: hb_dns_activity
verified: dry-run
verified_at: '2026-11-17'
~~~
SELECT device_hostname, query_hostname, answers, time FROM hb_dns_activity WHERE (('{{scope_hosts}}' = '' OR instr(',' || '{{scope_hosts}}' || ',', ',' || device_hostname || ',') > 0)) AND instr(',' || '{{tracking_domains}}' || ',', ',' || LOWER(query_hostname) || ',') > 0 AND time >= datetime('now', '-{{lookback_days}} days')
```
## rare-cloud-api-impact
<!-- Rare cloud API inbox rule manipulation -->
Detect account impact by finding users performing rare inbox rule operations or OAuth consents typical of BEC.
```sqlite target=endpoint role=detection-candidate params=(lookback_days=lookback_days)
~~~yaml
baseline:
compare: new_this_window
window: '{{lookback_days}}d'
expected: Specific users creating inbox rules when such behavior is rare across the
tenant. Silence suggests no automated rule manipulation has occurred.
prevalence:
by: actor_user_name
key:
- api_operation
rare_below: 4
reads:
- api_operation
- actor_user_name
- time
silence: not_evidence_of_absence
source: hb_cloud_api_activity
verified: dry-run
verified_at: '2026-11-17'
~~~
SELECT api_operation, actor_user_name, COUNT(DISTINCT actor_user_name) AS user_count, MIN(time) AS first_seen FROM hb_cloud_api_activity WHERE provider = 'm365' AND (api_operation IN ('New-InboxRule', 'Set-InboxRule', 'Consent to application')) AND time >= datetime('now', '-{{lookback_days}} days') GROUP BY api_operation, actor_user_name HAVING user_count <= 3
```
## triage-correlation
<!-- Correlate email, interaction, and abuse -->
```agent target=hunter
cite: required
context:
- verify-lead
- dns-tracking-interaction
- rare-cloud-api-impact
max_iterations: 5
objective: Evaluate the sequence of phishing delivery, tracking resolution, and cloud
API changes to determine if an account takeover occurred.
success_criteria: A verdict of malicious | suspicious | benign per user/host, citing
rows from all surfaces.
tools:
- endpoint
- microsoft-365
```
## route-on-triage
<!-- Route on compromise verdict -->
if~: "the triage-correlation verdict is malicious for at least one user" (confidence: high, judge=hunter)
then: → isolate-account
indeterminate: → analyst-review
unavailable: → analyst-review (blind_spot: no-api-retention)
else: → close-out
## isolate-account
<!-- Isolate compromised account -->
```action target=identity
~~~yaml
approval: required
~~~
Revoke all active refresh tokens for the identified users, reset their passwords, and initiate an audit of their mailbox for new rules.
```
→ analyst-review
## analyst-review
<!-- Forensic analyst review -->
```manual target=analyst
Examine the timing between DNS resolution of the tracking domains and the creation of inbox rules. Check SharePoint and Teams logs for unusual file access from the same actor IP.
```
→ close-out
## close-out
<!-- Close out hunt -->
```manual target=analyst
Record the findings. If inbox rule manipulation was confirmed, promote the rare-cloud-api-impact query to a standing detection.
```
→ end
Run it
Take this hunt into your environment.
Open it in Huntbase to run every step against your own connections, with Scout weighing the evidence and your analysts in command. Or take the open hunt.md file anywhere that reads the format.
Machine-drafted by huntbase-hunt-generation using hb_google/gemini-3-flash-preview, gated by dry-run, lint, critic, then reviewed by a person.