Microsoft Teams CVE-2026-42835

Exploitation requires the attacker to hold valid low-privileged credentials within a Teams context reachable by the victim (PR:L in the CVSS vector) - for example, a member of the same tenant, a federated external user, or a guest invited to a shared channel - and the victim must be running the vulnerable Microsoft Teams app on Android. No user interaction is required (UI:N), meaning the victim does not need to click or open the malicious content for the injection to trigger in the downstream component. Limiting factors include the requirement for some level of authenticated Teams access (preventing fully anonymous internet-wide exploitation), the Android-only client scope (Windows, macOS, iOS, and web Teams clients are not implicated by this CVE ID), and the specific tenant federation / guest-access policies that govern whether an external attacker can reach an internal victim at all.

The CVSS 3.1 vector (AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:H) indicates network-reachable exploitation with low attack complexity, requiring only low privileges (e.g., an authenticated Teams user) and no user interaction, with high confidentiality and availability impact but no integrity impact - an unusual combination suggesting data exfiltration plus a denial-of-service side effect. EPSS, KEV, and POC data were not provided in this intelligence package, so real-world exploitation likelihood cannot be quantified; absence of those signals combined with a Microsoft self-reported advisory suggests this was discovered internally or via coordinated disclosure rather than from in-the-wild observation. The 'Google' tag alongside Microsoft hints at joint discovery or an Android-platform interaction, but no further corroborating data is available - defenders should treat the 8.1 score as the primary signal and prioritize patching given mobile client ubiquity.

An attacker with a valid (possibly guest or federated) Teams account crafts a message, channel post, or other Teams content containing specially formatted payload characters that bypass output neutralization. When the downstream component on a victim's Android Teams client processes that content, the injection causes the client to disclose sensitive data - for example, leaking authentication tokens, message contents from other conversations, or device metadata - back to attacker-controlled infrastructure or visible channels, while also producing a high availability impact such as client crashes.