Threat intel5 min read
WhatsApp fixes three flaws, one possibly used to back up scam messages
WhatsApp's November advisories cover three bugs. A low-severity Android ad-reply flaw that leaks a phone's IP address may have been abused with spam. Update to 2.26.35.74 or later on Android.
4Tify Team

WhatsApp published three new security advisories on 20 November 2026. Two of them affect the Android app only, and one reaches the verifier the apps use to trust Meta AI's confidential computing backend. The one that matters most in practice is also the weakest on paper: a low-severity bug in Click to WhatsApp ad replies that WhatsApp says may have been abused together with specific spam messages.
None of the three is in the CISA Known Exploited Vulnerabilities catalog.
What happened
CVE-2026-23871 (CVSS 3.5, low). The Android apps did not fully validate the external URLs carried in ad-reply messages, the messages created when someone answers a Click to WhatsApp ad. An attacker could craft one of these messages so that, when it was shown in a chat, the phone loaded media from a server the attacker controls. That request exposes the phone's IP address.
WhatsApp's own security team found it. The company said the bug turned up while the team was disrupting a scam campaign, and that it believes a scammer used it to make scam messages look more credible. WhatsApp has not said who was behind the campaign, where it ran or how many people received the messages.
CVE-2026-102769 (CVSS 6.8, medium). A SQL injection in WhatsApp for Android, triggered when the recipient taps a specially crafted "order details" message sent from an unofficial WhatsApp client. It could delete WhatsApp files stored on the phone. If the user was online when the message arrived, an attacker could also learn a single bit of information about the chat database, by forcing a crash and timing the delivery receipt. An outside researcher reported it through Meta's bug bounty program, and WhatsApp says it has seen no exploitation.
CVE-2026-44908 (CVSS 4.8, medium). WhatsApp's remote-attestation verifier did not correctly follow a certificate's chain of trust when connecting to the confidential computing backend behind Meta AI features. Someone inside Meta's own infrastructure could have presented a VCEK certificate that was not signed by the hardware manufacturer's key and so bypassed the hardware root of trust. Any such certificate would first have to appear in the public transparency ledger used by WhatsApp Private Processing. WhatsApp says every certificate published there chains back to the hardware root, so the flaw was not used.
Who is affected
| CVE | Product | Fixed in |
|---|---|---|
| CVE-2026-23871 | WhatsApp and WhatsApp Business for Android | 2.26.17.4 |
| CVE-2026-102769 | WhatsApp for Android (also WhatsApp Business for Android in the CVE record) | 2.26.35.74 |
| CVE-2026-44908 | WhatsApp for iOS | 2.26.24.72 |
| CVE-2026-44908 | WhatsApp Business for iOS, WhatsApp for Mac | 2.26.23.72 |
| CVE-2026-44908 | WhatsApp for Android | 2.26.23.74 |
| CVE-2026-44908 | WhatsApp Business for Android | 2.26.23.70 |
An Android phone on 2.26.35.74 or later is covered for all three. The full wording of each entry is on WhatsApp's security advisories page, and the NVD record for the ad-reply bug is CVE-2026-23871.
Why a low-severity bug still matters
An IP address on its own gives away little. In a scam, though, it lets the sender confirm that a message was opened, roughly place the victim and tailor the next message. The CVSS score measures technical damage, and this is a reminder that scammers care about something else: anything that makes the next message more convincing.
The SQL injection is the more serious bug technically, but it needs a crafted message from an unofficial client and a tap from the victim, and there is no sign anyone used it.
What to do now
- Update WhatsApp on every device. On Android, make sure WhatsApp and WhatsApp Business are at 2.26.35.74 or later through Google Play. On iPhone and Mac, install the latest version from the App Store.
- Companies with managed phones should check the WhatsApp version through their mobile device management tool and push the update where it lags.
- Treat unexpected ad replies and "order details" messages with suspicion, especially from numbers you have not dealt with before. Don't tap through until you know who sent them.
- Report and block spam accounts from inside the chat, which also feeds WhatsApp's own anti-scam work.
Both Android bugs come down to the same thing: content from outside the app reaching a network request or a local database without enough checks. When our team runs a penetration test on a mobile app and the services behind it, that is the class of issue we go looking for, from URLs an attacker controls to message fields that end up inside a query.