How the attack works
Password spraying turns guessing around: instead of many passwords for one account, the attacker tries one or two common passwords for many accounts. Each account sees so few failures that its lockout never triggers. Ping's predefined password policies lock after 5 failed attempts for 15 minutes, and a sprayer stays below that for every user. What gives the attack away is the source: one address that fails for many different users in a short time.
PingOne records a failed password check as PASSWORD.CHECK_FAILED and keeps the address of the originating HTTP request in source.ipAddress. A sprayer who works from a guessed or outdated list of names also hits names that do not exist, and PingOne has its own activity for that case, PASSWORD.CHECK_FAILED_USER_NOT_FOUND, which Ping's event list describes as Password Check Failed due to User Not Found. A burst of it is a list being tested, whether the goal is to find valid names or to spray them. This page is derived from Ping's documentation of the events and their attributes: no public incident report describes spraying against PingOne at the audit event level.
Log source: the rules read PingOne audit activities. Include the Password event types in the PingOne webhook that feeds the SIEM, or poll the Audit Activities API, for every environment. Two of the rules group by source.ipAddress, so they need a SIEM integration that keeps that attribute.
What an attacker needs
- A list of user names, real or guessed, and access to the sign-on page of an application that uses PingOne.
- Accounts that sign on with a password alone, or a second factor the attacker can pass, for the spray to end in access.
Impact
Every account in the list that uses the tried password is compromised without a lock and without the user noticing. The attacker gains those users' sessions in every application that trusts PingOne, and the administrator console where one of the accounts holds a role.
Detection rules
Failed PingOne Password Checks for Many Users From One Source Address
Counts the different users for whom the password check failed (audit activity PASSWORD.CHECK_FAILED) from one source address. Ten or more users from one address within 30 minutes fit password spraying, where each account sees too few failures to lock. Ping documents source.ipAddress as the IP address of the originating HTTP request and lists it among the attributes of this activity. The users are counted by actors.user.name: Ping documents the actor as the end user responsible for the event, which in a sign-on flow is the user who is signing on. The count and the window are starting points to tune. A sprayer who paces the attempts or spreads them over many addresses stays below the count.
Fires when one source.ipAddress has at least 10 different actors.user.name values within 30 minutes.
title: PingOne Password Check Failed (Spraying Base Event)
id: bc8ec27a-a04f-44dc-8071-592b6e493ed4
name: pingone_password_check_failed_by_source
status: experimental
description: >-
Base event of the threshold rule in this file. The audit activity
PASSWORD.CHECK_FAILED, which Ping's event list describes as Password Check
Failed.
logsource:
product: pingone
service: audit
detection:
selection:
action.type: PASSWORD.CHECK_FAILED
condition: selection
level: low
---
title: Failed PingOne Password Checks for Many Users From One Source Address
id: 73c152a4-f918-431c-83b9-66ab0375d744
status: experimental
description: >-
Counts the different users for whom the password check failed (audit
activity PASSWORD.CHECK_FAILED) from one source address. Ten or more users
from one address within 30 minutes fit password spraying, where each
account sees too few failures to lock. Ping documents source.ipAddress as
the IP address of the originating HTTP request and lists it among the
attributes of this activity. The users are counted by actors.user.name:
Ping documents the actor as the end user responsible for the event, which
in a sign-on flow is the user who is signing on. The count and the window
are starting points to tune. A sprayer who paces the attempts or spreads
them over many addresses stays below the count.
references:
- https://attack.mitre.org/techniques/T1110/003/
- https://developer.pingidentity.com/pingone-api/platform/reference/audit-reporting-events.html
- https://developer.pingidentity.com/pingone-api/platform/audit-activities.html
author: Forestall
date: 2026-11-18
tags:
- attack.credential-access
- attack.t1110.003
correlation:
type: value_count
rules:
- pingone_password_check_failed_by_source
group-by:
- source.ipAddress
timespan: 30m
condition:
field: actors.user.name
gte: 10
falsepositives:
- Many people behind one office, VPN or proxy address who mistype their password in the same period, for example on the morning after a forced password change. Exclude your own egress addresses or give them a higher count.
- An application that calls the PingOne flow API from its own servers and therefore appears as the source address of all its users.
- The values of 10 users and 30 minutes are starting points, not tested limits. Measure what is normal for your environment and adjust both.
level: medium
False positives
- Many people behind one office, VPN or proxy address who mistype their password in the same period, for example on the morning after a forced password change. Exclude your own egress addresses or give them a higher count.
- An application that calls the PingOne flow API from its own servers and therefore appears as the source address of all its users.
- The values of 10 users and 30 minutes are starting points, not tested limits. Measure what is normal for your environment and adjust both.
Many PingOne Sign-On Attempts for Unknown Users Through One Client
Counts, per client, the password checks that failed because PingOne does not hold the user (audit activity PASSWORD.CHECK_FAILED_USER_NOT_FOUND). Ping documents the client actor as the client the actor used to perform the action, which in a sign-on flow is the application the user signs on to. Twenty or more through one client within 10 minutes mean that a list of names is being tested at that application's sign-on page, also when the attempts come from many source addresses and no single address stands out. The rule complements the per-address rule of this page and works in a SIEM integration that does not keep the source address. The count depends on how busy the application is, so it is a starting point to tune per client.
Fires when one actors.client.id has at least 20 matching events within 10 minutes.
title: PingOne Password Check Failed for an Unknown User (Per Client Base Event)
id: 98c6b0d7-d4d9-4cd9-acf5-183d3ea669ec
name: pingone_password_check_user_not_found_by_client
status: experimental
description: >-
Base event of the threshold rule in this file. The audit activity
PASSWORD.CHECK_FAILED_USER_NOT_FOUND, which Ping's event list describes as
Password Check Failed due to User Not Found.
logsource:
product: pingone
service: audit
detection:
selection:
action.type: PASSWORD.CHECK_FAILED_USER_NOT_FOUND
condition: selection
level: low
---
title: Many PingOne Sign-On Attempts for Unknown Users Through One Client
id: b5bd128f-d617-482f-818c-6f8aa319988c
status: experimental
description: >-
Counts, per client, the password checks that failed because PingOne does
not hold the user (audit activity PASSWORD.CHECK_FAILED_USER_NOT_FOUND).
Ping documents the client actor as the client the actor used to perform the
action, which in a sign-on flow is the application the user signs on to.
Twenty or more through one client within 10 minutes mean that a list of
names is being tested at that application's sign-on page, also when the
attempts come from many source addresses and no single address stands out.
The rule complements the per-address rule of this page and works in a SIEM
integration that does not keep the source address. The count depends on how
busy the application is, so it is a starting point to tune per client.
references:
- https://attack.mitre.org/techniques/T1110/003/
- https://developer.pingidentity.com/pingone-api/platform/reference/audit-reporting-events.html
- https://developer.pingidentity.com/pingone-api/platform/audit-activities.html
author: Forestall
date: 2026-11-18
tags:
- attack.credential-access
- attack.t1110.003
correlation:
type: event_count
rules:
- pingone_password_check_user_not_found_by_client
group-by:
- actors.client.id
timespan: 10m
condition:
gte: 20
falsepositives:
- A busy application whose users mistype their sign-on name, and the days after a migration that changed sign-on names. Raise the count for such clients or exclude them after review.
- A monitoring or load test that signs on with accounts that were deleted.
- The values of 20 attempts and 10 minutes are starting points, not tested limits.
level: low
False positives
- A busy application whose users mistype their sign-on name, and the days after a migration that changed sign-on names. Raise the count for such clients or exclude them after review.
- A monitoring or load test that signs on with accounts that were deleted.
- The values of 20 attempts and 10 minutes are starting points, not tested limits.
Many PingOne Sign-On Attempts for Unknown Users From One Source Address
Counts, per source address, the password checks that failed because PingOne does not hold the user (audit activity PASSWORD.CHECK_FAILED_USER_NOT_FOUND, described by Ping as Password Check Failed due to User Not Found). Ten or more from one address within 10 minutes mean that a list of names is being tested: either to learn which accounts exist, or as the misses of a password spray that runs on guessed or outdated names. Ping lists source.ipAddress, the IP address of the originating HTTP request, among the attributes of this activity. The rule counts activities, not names, because Ping does not document which attribute carries the name that was tried. The count and the window are starting points to tune.
Fires when one source.ipAddress has at least 10 matching events within 10 minutes.
title: PingOne Password Check Failed for an Unknown User (Per Address Base Event)
id: 229d4ef1-5c8c-4855-9523-bc47ff77da4d
name: pingone_password_check_user_not_found_by_source
status: experimental
description: >-
Base event of the threshold rule in this file. The audit activity
PASSWORD.CHECK_FAILED_USER_NOT_FOUND, which Ping's event list describes as
Password Check Failed due to User Not Found.
logsource:
product: pingone
service: audit
detection:
selection:
action.type: PASSWORD.CHECK_FAILED_USER_NOT_FOUND
condition: selection
level: low
---
title: Many PingOne Sign-On Attempts for Unknown Users From One Source Address
id: c7d257f6-06a5-4feb-b1c3-5d4e06436cf3
status: experimental
description: >-
Counts, per source address, the password checks that failed because PingOne
does not hold the user (audit activity PASSWORD.CHECK_FAILED_USER_NOT_FOUND,
described by Ping as Password Check Failed due to User Not Found). Ten or
more from one address within 10 minutes mean that a list of names is being
tested: either to learn which accounts exist, or as the misses of a
password spray that runs on guessed or outdated names. Ping lists
source.ipAddress, the IP address of the originating HTTP request, among the
attributes of this activity. The rule counts activities, not names, because
Ping does not document which attribute carries the name that was tried. The
count and the window are starting points to tune.
references:
- https://attack.mitre.org/techniques/T1110/003/
- https://developer.pingidentity.com/pingone-api/platform/reference/audit-reporting-events.html
- https://developer.pingidentity.com/pingone-api/platform/audit-activities.html
author: Forestall
date: 2026-11-18
tags:
- attack.credential-access
- attack.t1110.003
correlation:
type: event_count
rules:
- pingone_password_check_user_not_found_by_source
group-by:
- source.ipAddress
timespan: 10m
condition:
gte: 10
falsepositives:
- Users of a shared office, VPN or proxy address who try an old user name or the name of another directory, for example after a migration that changed the sign-on name.
- An application that calls the PingOne flow API from its own servers and therefore appears as the source address of all its users.
- The values of 10 attempts and 10 minutes are starting points, not tested limits.
level: medium
False positives
- Users of a shared office, VPN or proxy address who try an old user name or the name of another directory, for example after a migration that changed the sign-on name.
- An application that calls the PingOne flow API from its own servers and therefore appears as the source address of all its users.
- The values of 10 attempts and 10 minutes are starting points, not tested limits.
Detection logic
Patterns that need more than one event, a baseline or knowledge of your own environment. Each is a hypothesis to build and tune in your SIEM, not a ready rule.
Successful password check from an address that failed for many users
A spray matters when one of its attempts works. A successful password check from an address that, in the same period, failed for many other users or tried many unknown names is very likely the account whose password matched the sprayed one, and it should be handled as a compromised account.
- Build the set of addresses that the threshold rules of this page report:
source.ipAddressvalues with failed checks (PASSWORD.CHECK_FAILED) for manyactors.user.namevalues, or with manyPASSWORD.CHECK_FAILED_USER_NOT_FOUNDactivities. - Look for
PASSWORD.CHECK_SUCCEEDEDactivities with the samesource.ipAddressfrom one hour before the first failure to 24 hours after the last. - For each such user, check whether an MFA step followed and passed, and list the activities of the user after the sign-on.
EventsPASSWORD.CHECK_FAILEDPASSWORD.CHECK_FAILED_USER_NOT_FOUNDPASSWORD.CHECK_SUCCEEDED
LimitsUsers behind a shared office or VPN address sign on successfully from it all day, so exclude your own egress addresses first. A sprayer who spreads the attempts over many addresses never enters the set, and the time spans are starting points to tune.
Triage
- List the user names the address failed for (actors.user.name) and how the attempts are spread over time. An even rhythm with one attempt per user is a spray, while a few users with several attempts each is more likely a shared office address.
- Identify the address. An egress address of your own offices, VPN or identity-aware proxy explains many users, a hosting provider or anonymizing network does not.
- Look for
PASSWORD.CHECK_SUCCEEDEDactivities from the same address in the same period. Each one is an account whose password the source knows. - For unknown user names, compare the tried names with your naming scheme and with former employees. Names that follow your scheme mean the attacker knows it.
- Check the client (actors.client.name) to see which application's sign-on page is being used, and whether that application requires MFA.
Response
Once the attack is confirmed:
- Block the source addresses at the network edge or with a sign-on policy condition, and watch for the same pattern from new addresses.
- For every account that signed on from the source, end the sessions, reset the password and review the MFA devices registered since the first attempt.
- Force a password change for the accounts the source tried if you can identify the password that was sprayed, for example from a successful attempt.
- Require MFA on the application whose sign-on page was used, if it accepted a password alone.
- Review what the compromised accounts did after sign-on, above all role assignments, new applications, and changes to their own MFA devices.
Mitigation
- Require MFA in the sign-on policy or flow of every application, so that a correct password from a spray is not enough.
- Enable the check against commonly used passwords in every password policy (
excludesCommonlyUsed), which removes the passwords a spray tries first. - Keep a lockout rule in every password policy. It does not stop a slow spray, but it limits what one address can try per account.
- Put a risk policy or a bot detection step in front of the password prompt of applications that face the internet, and block sources that fail for many accounts.
- Send the Password event types to the SIEM with the source address, for every environment.
Frequently asked questions
Why does the account lockout not stop password spraying?
The lockout counts failed attempts per account. Ping's predefined password policies lock after 5 failed attempts for 15 minutes, and a spray tries only one or two passwords per account and then moves to the next. No account reaches its limit. The pattern only shows when failures are counted per source address across accounts, which is what the first rule does.
What does PASSWORD.CHECK_FAILED_USER_NOT_FOUND tell a defender?
Ping's event list describes it as Password Check Failed due to User Not Found: someone tried to sign on with a name PingOne does not hold. A few a day are typing mistakes. Many in a short time from one address, or at one application, mean that a list of names is being tested, either to find out which accounts exist or as the misses of a spray that runs on guessed or outdated names.
Why do two of the rules need the source address?
Spraying is defined by its source: many accounts from one place. Ping documents `source.ipAddress` as the IP address of the originating HTTP request for these activities. A SIEM integration that does not keep the attribute cannot run the per-address rules. The third rule counts unknown user names per client instead, which also catches a list that is tested from many addresses.
Is the source address always the user's own address?
Not always. Ping documents it as the address of the originating HTTP request. When users sign on through PingOne's hosted pages, that is the browser's address. An application that calls the flow API from its own servers, or a proxy in front of sign-on, can appear as the source for all of its users. Identify such addresses first and exclude them or count them separately.