Skip to content

WhatsApp Patches 3 Flaws, Including an Android Scam Bug

WhatsApp Patches 3 Flaws, Including an Android Bug Suspected in Scams

One of the three bugs may have helped a scammer. The other two have no reported abuse.

WhatsApp Patches 3 Flaws, Including an Android Scam Bug

WhatsApp fixed three vulnerabilities in its November 2026 security advisory. Only one, CVE-2026-23871 in the Android apps, comes with suspected abuse: WhatsApp believes a scammer used it to make a scam message look more believable. Update WhatsApp and WhatsApp Business on Android, because the affected builds for that flaw are those before 2.26.17.4. [1][2]

  • The advisory covers CVE-2026-23871, CVE-2026-102769 and CVE-2026-44908.
  • Only CVE-2026-23871 has suspected abuse, tied to specific spam messages.
  • The advisory describes that flaw as exposing a target's IP information, nothing more.
  • WhatsApp reports no evidence of exploitation for the other two.
  • No CVSS scores turned up in the sources we opened, and we did not check CISA KEV status.

Most scammers never touch a vulnerability. They send a convincing message and wait. WhatsApp's advisory describes a rare exception, a bug that a scammer may have used to polish a lure.

This is a different problem from the zero-click spyware attacks on messaging apps we have covered. Nobody is reported to have lost control of a phone here. The human side of the story sits in our guide to social engineering defenses.

If you rank bugs by score, our notes on CVSS 4.0 triage explain how, once a score is published for these.

What did WhatsApp actually fix?

WhatsApp fixed three separate vulnerabilities in its November 2026 security advisory. CVE-2026-23871 is an Android ad-reply flaw that could expose a target's IP information. CVE-2026-102769 is an Android SQL injection bug that could delete local WhatsApp files. CVE-2026-44908 is a certificate trust flaw in the check WhatsApp apps use to verify Meta AI's confidential computing backend. WhatsApp lists each one separately. [1][3]

The graphic below sorts the three WhatsApp flaws by what each one touched and what is known about abuse.

The three WhatsApp flaws in the November 2026 advisory CVE-2026-23871 is an Android ad-reply IP exposure with suspected abuse alongside spam messages. CVE-2026-102769 is an Android SQL injection with no evidence of exploitation. CVE-2026-44908 is a Meta AI attestation trust flaw where the transparency ledger shows no exploitation. CVE-2026-23871 Android ad-reply, IP exposure Suspected abuse with spam messages CVE-2026-102769 Android SQL injection No evidence of exploitation CVE-2026-44908 Meta AI attestation trust check Ledger shows no exploitation
Overview of the three flaws in WhatsApp's November 2026 advisory. Source: WhatsApp Security Advisories [1]. Conceptual summary graphic by UnpanicTech.

How did the Android ad-reply flaw (CVE-2026-23871) work?

CVE-2026-23871 comes from incomplete validation of external ad-reply URLs in WhatsApp for Android and WhatsApp Business for Android before version 2.26.17.4. If an attacker sent a crafted Click to WhatsApp ad-reply message with media hosted on a server they controlled, the target's device could reveal its IP information when the message was shown in a chat. [1]

Click to WhatsApp ads are ads that open a chat with a business when tapped. The advisory describes the attack as a crafted ad-reply message, which is the message that appears after someone responds to such an ad. [1]

Sequence described for CVE-2026-23871 Step one, an attacker sends a crafted ad-reply message. Step two, the media points to a server the attacker controls. Step three, the message renders in the chat thread. Step four, the target's device reveals its IP information. 1. Crafted ad-reply message sent by an attacker 2. Media points to a server that the attacker controls 3. Message renders in the chat thread on the target's phone 4. Target's device reveals its IP information
Conceptual sequence based on WhatsApp's description of CVE-2026-23871. This is an illustration, not a screenshot or a capture of an attack.

An IP address usually points to a rough location and a network provider. That is general background, not something WhatsApp said about this case. The advisory mentions IP information and nothing else.

The bug was found by internal researchers on the WhatsApp Security Team. A bug bounty report did not trigger it. [1]

Was CVE-2026-23871 actually exploited?

WhatsApp says CVE-2026-23871 may have been abused in conjunction with specific spam messages. A company spokesperson told The Hacker News and SecurityWeek that the team found the flaw while disrupting a scam campaign and believes a scammer used it to make a message seem more credible. WhatsApp has not disclosed how many people were reached or whether the scam worked. [1][2][3]

Wording matters here. The advisory says "may have been abused." The spokesperson says "we believe." Security Affairs goes further in its headline and summary, saying the bug was exploited in a real scam campaign [4]. That is a fair reading of the vendor's belief, but it is firmer than the vendor's own text.

Here is how the evidence breaks down:

  • Confirmed by WhatsApp: the flaw existed and is fixed.
  • Vendor belief: a scammer used it to add credibility to a message.
  • Not public: the scale, the success rate, and how exactly the exposure made a message look better.

One reading, ours and not WhatsApp's: attacker-chosen media shown inside an ad-reply message could make a lure look like a legitimate business exchange. The sources we opened do not confirm that.

What about the other two flaws, CVE-2026-102769 and CVE-2026-44908?

CVE-2026-102769 is an Android SQL injection bug triggered by tapping a crafted order details message from an unofficial client. It could delete local WhatsApp files and, in narrow conditions, leak one bit about the chat database. CVE-2026-44908 is a trust-validation flaw that only someone positioned inside Meta infrastructure could have used. WhatsApp reports no evidence of exploitation for either one. [1]

The CVE-2026-102769 data leak needed the user to be online when the message arrived. WhatsApp describes it as one bit of information, chosen by the attacker, about the chat database, learned through an induced-crash delivery receipt timing side channel. In plain terms, that is a single yes or no answer. [1]

CVE-2026-44908 sits in the check WhatsApp apps use to confirm they can trust Meta AI's confidential computing backend. Someone inside Meta infrastructure could have presented a VCEK certificate signed by the wrong key. That certificate would first have had to appear on a third-party transparency ledger, and WhatsApp says the ledger shows every published certificate chains to the hardware root of trust. [1]

The press disagrees on which flaw matters most. SecurityWeek calls CVE-2026-23871 low-impact and the other two medium-impact. Security Affairs says the SQL injection is more serious on paper, yet only the ad-reply flaw saw reported abuse. Those labels are editorial judgments, not CVSS ratings. [3] [4]

Which WhatsApp versions are affected?

The fixed-version cutoffs differ by flaw and platform. By our reading of the advisory, WhatsApp for Android needs to be at 2.26.35.74 or later to clear all three items, and WhatsApp Business for Android needs 2.26.23.70 or later for the items that list it. iOS and Mac builds appear only under CVE-2026-44908. [1]

The table lays out the WhatsApp advisory's version data by flaw and app. Every row means "affected before this version."

Flaw App Affected before Abuse status per WhatsApp
CVE-2026-23871 WhatsApp and WhatsApp Business for Android 2.26.17.4 May have been abused with spam messages
CVE-2026-102769 WhatsApp for Android 2.26.35.74 No evidence of exploitation
CVE-2026-44908 WhatsApp for iOS 2.26.24.72 Ledger shows no exploitation
CVE-2026-44908 WhatsApp Business for iOS and WhatsApp for Mac Desktop 2.26.23.72 Ledger shows no exploitation
CVE-2026-44908 WhatsApp for Android 2.26.23.74 Ledger shows no exploitation
CVE-2026-44908 WhatsApp Business for Android 2.26.23.70 Ledger shows no exploitation

Source: WhatsApp Security Advisories 2026, November update [1]. The advisory does not list WhatsApp Business for Android under CVE-2026-102769.

What should you do now?

Update WhatsApp and WhatsApp Business from your device's official app store, then confirm the installed version is past the cutoffs in the table. There is no setting to change for the ad-reply flaw. WhatsApp's own advice is to keep apps and devices up to date. [1][2]

  1. Open Google Play or the App Store and update WhatsApp, plus WhatsApp Business if you use it.
  2. Check the version afterwards. In WhatsApp on Android this is usually under Settings, Help, App info.
  3. Treat unexpected ad-reply or order details messages as unverified. Check the business through its official website, not through the chat.
  4. Ignore urgency. A message that pushes you to act in minutes deserves less trust, not more.

Steps three and four are UnpanicTech's advice. They do not come from the advisory.

What is still unknown about these WhatsApp flaws?

WhatsApp has not published how many people received messages tied to CVE-2026-23871, whether any scam succeeded, or how the IP exposure made a message look more believable. The sources UnpanicTech opened give no CVSS scores for any of the three flaws. CISA Known Exploited Vulnerabilities status was not checked, so nobody should read this article as saying it is or is not listed. [1][2]

We did not reproduce any of these bugs. Everything here comes from WhatsApp's advisory and the press coverage cited below.

What does this mean for defenders and ordinary users?

The WhatsApp ad-reply flaw had the lowest technical impact of the three and still earned the only abuse suspicion. A flaw does not need to be severe to be useful to a scammer. It only needs to make a lure look a little more real.

For teams, the lesson is about inventory. Staff run WhatsApp and WhatsApp Business on personal and company phones, and the advisory's cutoffs differ per app and per platform. A ticket that says "update WhatsApp" is too loose. List the app, the platform and the version.

Anyone who runs Click to WhatsApp ads should expect customers to see lookalike ad-reply messages, and should tell them where genuine contact comes from. The ad-reply flaw gave no reported access to a device, so the defense that matters most is human: slow down, verify, and do not treat polish as proof.

WhatsApp's wording may also change as it learns more, since the company called the abuse suspected, not confirmed. Re-read the advisory page in a few weeks and check whether the description of CVE-2026-23871 has moved.

Sources & References

  1. WhatsApp, "WhatsApp Security Advisories 2026" (November update): https://www.whatsapp.com/security/advisories/2026/
  2. The Hacker News, "WhatsApp Patches 3 Security Flaws, Including Android Bug Suspected of Use in Scams," November 20, 2026: https://thehackernews.com/2026/11/whatsapp-patches-3-security-flaws.html
  3. SecurityWeek, "WhatsApp Fixes Flaw Uncovered While Disrupting Scam Campaign," November 20, 2026: https://www.securityweek.com/whatsapp-fixes-flaw-uncovered-while-disrupting-scam-campaign/
  4. Security Affairs, "CVE-2026-23871: WhatsApp Bug Helped Scammers Make Messages Look Legitimate," November 20, 2026: https://securityaffairs.com/203622/security/cve-2026-23871-whatsapp-bug-helped-scammers-make-messages-look-legitimate.html
NK

Cybersecurity Researcher & Technical Editor

UnpanicTech publishes independent cybersecurity research and defensive security analysis. Naseem Khan serves as Technical Editor, with a focus on vulnerability analysis, defensive security, incident response, cloud security, and practical security engineering.

Leave a Comment (Authenticated Users)