WhatsApp CVE-2026-23871, CVE-2026-102769: Update Guide – ThreatDefence
Skip to content
Security advisories/Meta (WhatsApp)
MediumActive

WhatsApp CVE-2026-23871 and CVE-2026-102769: Patching Guide

Roche Zhang
Roche Zhang · Senior Security Analyst
Published November 21, 2026

Meta fixed WhatsApp flaws in November 2026. The one to act on, CVE-2026-23871, lets a crafted ad reply reveal an Android user’s IP address, and Meta says scammers may have used it.

TL;DR

Update WhatsApp on Android to 2.26.35.74 or later, and on iPhone and Mac to the November builds.

At a glance
SeverityMedium
StatusActive
ExploitedMeta says CVE-2026-23871 may have been abused in scam messages to reveal recipients' IP addresses. No exploitation seen for the other two. None is in CISA KEV
ProductWhatsApp and WhatsApp Business for Android, WhatsApp and WhatsApp Business for iOS, and WhatsApp for Mac. WhatsApp for Windows is not affected
CVEsCVE-2026-23871, CVE-2026-102769, CVE-2026-44908
Fixed inAndroid needs WhatsApp 2.26.35.74 or later, which covers every fix. iOS needs 2.26.24.72 (Business 2.26.23.72) and Mac 2.26.23.72. Meta published the fixes in its November 2026 update, with CVE records on November 20

Meta’s November 2026 WhatsApp security update fixes flaws in the Android, iOS and Mac apps, with CVE records published on November 20, 2026. None is a remote takeover of the phone, and none is in CISA’s Known Exploited Vulnerabilities catalog. One has been used in the real world, though:

  • CVE-2026-23871, in WhatsApp and WhatsApp Business for Android, lets a crafted Click to WhatsApp ad reply make the recipient’s phone fetch media from a server the sender controls, which reveals the phone’s IP address. Meta says it “may have been abused in conjunction with specific spam messages”, and a WhatsApp spokesperson said a scammer used it to make scam messages look more credible.
  • CVE-2026-102769, a SQL injection in WhatsApp for Android, triggered by a crafted “order details” message sent from a modified client. It can delete WhatsApp files on the phone and leak one bit of information about the chat database at a time. Meta has not seen it exploited.
  • CVE-2026-44908, a flaw in how WhatsApp checks the hardware attestation of Meta AI Private Processing, the confidential computing service behind some Meta AI features. Only someone inside Meta’s infrastructure could use it, and Meta says its public transparency log confirms it was not exploited.

WhatsApp for Windows is not part of this update. WhatsApp for Mac is affected only by CVE-2026-44908.

CVE CVSS EPSS CISA KEV Fixed in
CVE-2026-23871 3.5 (CVSS 3.1), Meta Not yet scored No WhatsApp and WhatsApp Business for Android 2.26.17.4
CVE-2026-102769 6.8 (CVSS 3.1), Meta Not yet scored No WhatsApp for Android 2.26.35.74 (Meta’s data also lists WhatsApp Business for Android)
CVE-2026-44908 4.8 (CVSS 3.1), Meta Not yet scored No WhatsApp for iOS 2.26.24.72, WhatsApp Business for iOS and WhatsApp for Mac 2.26.23.72, WhatsApp for Android 2.26.23.74, WhatsApp Business for Android 2.26.23.70

CVSS and EPSS from NVD and FIRST, exploitation status from the CISA Known Exploited Vulnerabilities catalog, checked November 21, 2026.

What is confirmed

Point Status
CVE-2026-23871, CVE-2026-102769 and CVE-2026-44908 fixed in the November 2026 update, CVE records published November 20 Meta
CVE-2026-23871 may have been abused in spam and scam messages to reveal IP addresses Meta, and a WhatsApp spokesperson
No exploitation seen for CVE-2026-102769 Meta
No exploitation of CVE-2026-44908, confirmed by the third-party transparency log Meta
CVSS 3.1: 3.5, 6.8 and 4.8 Meta. NVD had not scored them on November 21
CISA KEV None listed. CISA’s own rating on the CVE records says exploitation “none”
Public exploit None published. Meta’s score for CVE-2026-102769 says functional exploit code exists, which reflects the bug bounty report
WhatsApp for Windows Not affected

How the flaws work

CVE-2026-23871: IP address leak through ad replies. Businesses use Click to WhatsApp ads to start a chat from Facebook or Instagram. When someone replies to such an ad, the chat shows a preview of the ad. Meta says WhatsApp did not fully check the external URLs in these ad replies, so an attacker could send a crafted ad-reply message whose media came from their own server. When the chat displayed it, the phone fetched that media directly and revealed its IP address. The attacker needs a WhatsApp account, and the target has to open the chat. It affects WhatsApp and WhatsApp Business for Android from 2.20.198.3 up to, but not including, 2.26.17.4.

An IP address tells a scammer the target’s rough location and internet provider, which helps make a follow-up message believable (“we noticed a login from Sydney”). It can also confirm that a number is active and that the person opened the message.

CVE-2026-102769: SQL injection through order messages. WhatsApp Business supports structured “order details” messages. A crafted one, sent from an unofficial modified WhatsApp client, could inject SQL into the app’s local database queries when the recipient tapped it. Meta says this could delete WhatsApp files stored on the phone. If the recipient was online, the attacker could also force a crash and use the timing of delivery receipts to learn one bit of information of their choice about the chat database. Meta’s description names WhatsApp for Android before 2.26.35.74, and its structured CVE data also lists WhatsApp Business for Android from 2.22.3.7.

CVE-2026-44908: Meta AI attestation check. Private Processing runs some Meta AI requests inside confidential computing hardware, and the app checks a certificate chain to confirm it is talking to genuine hardware. WhatsApp did not follow that chain all the way to the hardware maker’s root. A party able to present a fake hardware certificate, which in practice means someone inside Meta’s infrastructure, could have bypassed the check. Meta says any such certificate would have had to appear on a public transparency log, and every certificate there chains correctly.

Practical exploit effort

The scores are low to medium, and the realistic risk matches them.

CVE-2026-23871 is easy to use and already used, with a narrow payoff. An attacker needs only a WhatsApp account, a crafted ad-reply message and a web server to log requests. No exploit code has been published, but the technique is simple enough that scammers used it before the fix. It reveals an IP address and that the message was opened. It does not give access to the phone, messages or account.

CVE-2026-102769 takes more work and gives limited access. The attacker needs a modified WhatsApp client to send the crafted order message, and the target has to tap it. Deleting WhatsApp files is the clearest impact. Reading data is slow: one bit per attempt, through forced crashes and timing, while the target is online. Repeated crashes of WhatsApp after tapping an order message would be the visible sign.

CVE-2026-44908 is out of reach for outside attackers. It needs control of Meta’s backend or the ability to issue hardware certificates.

Our assessment of the effort, as of November 21:

CVE-2026-23871 CVE-2026-102769 CVE-2026-44908
Public exploit No, but used by scammers No No
Attacker needs A WhatsApp account and a web server A modified WhatsApp client and a tap from the target A position inside Meta’s infrastructure
What it gives The target’s IP address and proof the chat was opened Deleted WhatsApp files, slow leaks from the chat database A bypass of the Meta AI hardware attestation check
Most exposed Android users who receive messages from unknown businesses or numbers Android users of WhatsApp Business features Users of Meta AI features in WhatsApp
Visible sign Unexpected ad-reply previews from unknown senders WhatsApp crashing after an order message, missing media None for users

For most organizations this is a routine app update with one social engineering angle. Staff who handle money, executives and anyone targeted by scams are the people for whom a leaked IP address matters most.

Who is affected, in priority order

  1. Staff on Android phones running WhatsApp or WhatsApp Business below 2.26.35.74, which covers every Android fix.
  2. Executives, finance and help desk staff who are frequent scam targets, for CVE-2026-23871.
  3. Businesses that use WhatsApp Business for orders and customer chats, for CVE-2026-102769.
  4. iPhone and Mac users, for CVE-2026-44908 only, with little practical risk.

Affected and fixed versions

App CVE-2026-23871 CVE-2026-102769 CVE-2026-44908 Install at least
WhatsApp for Android Fixed in 2.26.17.4 Fixed in 2.26.35.74 Fixed in 2.26.23.74 2.26.35.74
WhatsApp Business for Android Fixed in 2.26.17.4 Listed in Meta’s data, fixed in 2.26.35.74 Fixed in 2.26.23.70 2.26.35.74
WhatsApp for iOS Not affected Not affected Fixed in 2.26.24.72 2.26.24.72
WhatsApp Business for iOS Not affected Not affected Fixed in 2.26.23.72 2.26.23.72
WhatsApp for Mac Not affected Not affected Fixed in 2.26.23.72 2.26.23.72
WhatsApp for Windows Not affected Not affected Not affected No action

How to check your exposure

On a phone, the WhatsApp version is under Settings, Help (on Android, then App info). On a Mac it is under WhatsApp, About WhatsApp.

Across a fleet, use your mobile device management inventory. WhatsApp does not include the Microsoft Intune App SDK, so app protection policies cannot enforce a minimum version. Use the discovered apps report in Intune, or the equivalent in your MDM, to list WhatsApp and WhatsApp Business versions on managed Android and iOS devices, and push updates through managed Google Play or Apple Volume Purchase Program apps.

On personal phones enrolled only for work profiles or app protection, you will not see the personal WhatsApp version. Ask staff to update directly, starting with the groups above.

Mitigation and remediation

  1. Update WhatsApp on Android to 2.26.35.74 or later, on iPhone to 2.26.24.72 or later (Business 2.26.23.72), and on Mac to 2.26.23.72 or later. Turn on automatic updates in Google Play and the App Store.
  2. Use MDM to enforce updates on managed devices, and report on devices that stay on older versions.
  3. Tell staff about the scam angle. A message that shows an ad preview from an unknown business, or that quotes the recipient’s city or internet provider, is a reason for more caution, not less.
  4. Limit who can reach staff on WhatsApp. In Settings, Privacy, Advanced, the “Protect IP address in calls” setting relays calls through WhatsApp’s servers, and “Silence unknown callers” reduces cold contact. Neither fixes CVE-2026-23871, which concerns messages, but both reduce what scammers learn.
  5. Move sensitive business conversations to approved, managed channels where you can.
Free SOC trial

Want this watched across your environment without building the hunts yourself? Our 24x7 SOC monitors your endpoints, servers and cloud during a free trial.

Start a free trial

Threat hunting

No indicators of compromise have been published for this update, and the activity happens on phones, where most organizations have little telemetry. Hunting focuses on finding old versions and on the scam pattern.

Example KQL for Microsoft Defender and Microsoft Sentinel

The KQL below is example code for Microsoft Defender Vulnerability Management and Microsoft Sentinel. Adapt table and field names to your environment and test each query before relying on it.

Hunt 1: WhatsApp versions on onboarded devices

// Example: Defender Vulnerability Management inventory, WhatsApp by platform and version
DeviceTvmSoftwareInventory
| where SoftwareName has "whatsapp"
| summarize Devices=dcount(DeviceId), DeviceList=make_set(DeviceName, 50)
    by OSPlatform, SoftwareVendor, SoftwareName, SoftwareVersion
| order by OSPlatform asc, SoftwareVersion asc

This covers Mac and Windows devices, and Android and iOS devices onboarded to Defender for Endpoint. Compare each version with the table above.

Hunt 2: phishing and scam reports that mention WhatsApp

// Example: Sentinel, user-reported messages and alerts that mention WhatsApp in the last 30 days
SecurityAlert
| where TimeGenerated > ago(30d)
| where AlertName has "whatsapp" or Description has "whatsapp" or Entities has "whatsapp"
| project TimeGenerated, AlertName, AlertSeverity, ProviderName, Description
| order by TimeGenerated desc

Scams that start on WhatsApp often move to email or a phone call. Look for help desk tickets and reported phishing that mention WhatsApp contact, a business ad or the recipient’s location.

Responding to a suspected incident

  1. Update WhatsApp on the affected phone.
  2. Treat the sender as hostile. Block and report the number in WhatsApp, and warn colleagues who may receive the same message.
  3. Assume the attacker knows the target’s rough location and provider, and that the number is active. Brief the person on likely follow-up scams, such as fake account alerts or payment requests.
  4. For CVE-2026-102769, if WhatsApp crashed after an order message or media went missing, back up the chats, update and report the sender to WhatsApp.

Timeline

  • April 7, 2026: Trail of Bits publishes its audit of WhatsApp Private Processing before launch.
  • November 2026: Meta publishes its November WhatsApp security update.
  • November 20, 2026: Meta, as the CVE authority for its products, publishes CVE-2026-23871, CVE-2026-102769 and CVE-2026-44908. A WhatsApp spokesperson says CVE-2026-23871 was found while disrupting a scam campaign. CISA rates exploitation of each as none on the CVE records.
  • November 21, 2026: None of the CVEs is in CISA KEV, and NVD has not scored them.

Frequently Asked Questions

Is CVE-2026-23871 being exploited? Meta says it may have been abused with specific spam messages, and a WhatsApp spokesperson said a scammer used it to make scam messages more credible. It reveals the recipient’s IP address, not their messages. It is not in CISA KEV.

Which WhatsApp version fixes these flaws? On Android, 2.26.35.74 or later covers every fix. On iPhone, 2.26.24.72 or later, or 2.26.23.72 for WhatsApp Business. On Mac, 2.26.23.72 or later.

Can these flaws take over a phone or a WhatsApp account? No. CVE-2026-23871 leaks an IP address, CVE-2026-102769 can delete WhatsApp files and leak small amounts of database information, and CVE-2026-44908 needs access to Meta’s own infrastructure.

Is WhatsApp for Windows affected? No. None of these flaws affects the Windows app.

Is this a zero-click flaw? No. CVE-2026-23871 needs the target to open the chat, and CVE-2026-102769 needs the target to tap the order message.

What can someone do with my IP address? They learn your approximate location and internet provider, and that you opened their message. Scammers use this to make follow-up messages more convincing.

Should companies block WhatsApp? Not for this update. Make sure managed devices update, warn high-risk staff about the scam pattern, and keep sensitive business conversations in approved channels.

How ThreatDefence is responding

ThreatDefence has added CVE-2026-23871, CVE-2026-102769 and CVE-2026-44908 to our vulnerability intelligence. Across all monitored environments, our SOC is:

  • identifying devices with WhatsApp below the fixed versions, where mobile and endpoint inventory is available
  • watching for scam and phishing reports that start with WhatsApp contact, and helping customers brief high-risk staff
  • helping customers enforce WhatsApp updates through their device management

If you want help checking app versions across your fleet or investigating a suspicious message, contact us. Read more about our managed detection and response.

Need help checking your environment for this issue? Talk to the ThreatDefence SOC.Talk to Our SOC

Talk to Our SOC

Questions about this advisory or help checking your environment.

    I'm an MSP, MSSP, or reseller

    By submitting this form, you accept the terms in our privacy policy.