Skip to content
Security & Trust

WhatsApp believes a scammer used a flaw SecurityWeek calls low impact in a scam

SecurityWeek describes CVE-2026-23871 as low impact. WhatsApp's team found it while disrupting a scam campaign.

W
WebPulse Newsroom
AI-assisted · 4 min read
Share on X LinkedIn
WhatsApp believes a scammer used a flaw SecurityWeek calls low impact in a scam
In brief
  • WhatsApp believes a scammer used CVE-2026-23871, a bug SecurityWeek calls low impact, to make a scam message seem more credible. It is fixed in Android 2.26.17.4.
  • The flaw let a crafted ad-reply message make a phone reveal its IP information. Two other fixes, CVE-2026-102769 and CVE-2026-44908, show no evidence of exploitation, per WhatsApp.
  • Do not skip fixes because of a low rating. Check which WhatsApp versions staff run and how staff are told to treat unexpected business messages.

A low rating does not measure what a scammer gains

A severity rating answers one question: how much harm can this flaw do to a device? A scammer asks a different one: will this make my message believable? WhatsApp's latest security advisories show how far apart those questions can be.

WhatsApp's security team found the flaw while disrupting a scam campaign. A spokesperson told SecurityWeek: "we believe a scammer was using it to make their scam message seem more credible." SecurityWeek describes the bug, CVE-2026-23871, as low impact. The advisory itself says it "may have been abused in conjunction with specific spam messages."

How the flaw works

Click to WhatsApp ads are ads that open a chat with the advertiser. When one is answered, WhatsApp shows an ad-reply message in the thread. That message can pull in media from an outside web address.

WhatsApp says its Android apps did not fully check that address. An attacker could point it at a server they controlled. When the crafted message appeared in a chat, the target's phone could reveal its IP information to the attacker. The fix is in WhatsApp and WhatsApp Business for Android from version 2.26.17.4.

The sources leave gaps. WhatsApp does not explain how this leak made a scam more convincing. It also does not say how many people were targeted. The credibility claim is WhatsApp's belief, not a proven finding.

Two further fixes, with no exploitation seen

WhatsApp published two more advisories, which SecurityWeek rates medium impact. For both, WhatsApp says it has seen no evidence of exploitation.

The first, CVE-2026-102769, is a database-injection flaw in WhatsApp for Android builds older than 2.26.35.74. In plain terms, the app's database treats attacker-written text as commands. WhatsApp says a recipient who tapped a crafted "order details" message from an unofficial client could lose locally stored WhatsApp files. If the recipient was online, the attacker could also learn a single yes-or-no answer about the chat database. WhatsApp says this relied on the timing of a delivery receipt after an induced crash. An external researcher reported it through Meta's bug bounty.

The second, CVE-2026-44908, concerns the check WhatsApp apps run before trusting Meta AI's confidential computing backend. WhatsApp says a party inside Meta's own infrastructure could have defeated that check. The trick would be to show apps a forged-style certificate signed with the wrong key, not the chip maker's. WhatsApp adds that any such certificate would first have to be published to a third-party transparency ledger. Its statement: "The ledger shows that every published VCEK certificate chains to the hardware root of trust, confirming no exploitation." Meta's own researchers found this one.

The same advisory page also lists a Windows attachment-spoofing fix and a fix for AI rich response messages. WhatsApp reports no evidence of exploitation for either. The source gives no CVE IDs for them.

The lesson: judge a flaw by how it will be used

This shows a gap between technical severity and business risk. A bug that cannot take over a phone can still be useful to someone whose product is trust. Scam messages succeed when they look like the real thing. Anything that helps one look real has value to the scammer, whatever its score.

One case does not make a trend. It does show why a rating should be where the triage conversation starts, not where it ends.

What to ask your team

First, which WhatsApp and WhatsApp Business versions do staff run on Android, and who is responsible for updates? Companies that use WhatsApp for customer contact should confirm devices are at or beyond 2.26.17.4 for this fix. The order-details fix needs 2.26.35.74.

Second, does your patch policy let "low" items wait indefinitely? Ask for a short written reason when one is deferred.

Third, what do staff and customers do with an unexpected ad-reply or order message? A message that looks authentic is not proof that it is. Give people a simple way to report suspect messages and a second channel to verify them.

A flaw can be small and still do the scammer's work for them.

Produced by the WebPulse Newsroom with AI assistance from the original reporting credited below, and checked against that source by our editorial review. How we use AI.
Original reporting: WhatsApp (Meta).