CVE-2026-26133
published 2026-03-16CVE-2026-26133: AI command injection in M365 Copilot allows an unauthorized attacker to disclose information over a network.
PriorityP342high7.1CVSS 3.1
AVNACLPRNUIRSUCHILAN
EPSS
0.43%
35.1th percentile
AI command injection in M365 Copilot allows an unauthorized attacker to disclose information over a network.
Affected
55 ranges· showing 25
| Vendor | Product | Version range | Fixed in |
|---|---|---|---|
| microsoft | 365_copilot | < 2.107.2 | 2.107.2 |
| microsoft | 365_copilot | < 16.0.19815.10000 | 16.0.19815.10000 |
| microsoft | edge | < 145.3800.99 | 145.3800.99 |
| microsoft | excel | < 2.106.2 | 2.106.2 |
| microsoft | excel | < 16.0.19822.20038 | 16.0.19822.20038 |
| microsoft | loop | < 2.106 | 2.106 |
| microsoft | microsoft_365_copilot_for_android | >= 1.0 < 16.0.19815.10000 | 16.0.19815.10000 |
| microsoft | microsoft_365_copilot_for_ios | >= 1.0 < 2.107.2 | 2.107.2 |
| microsoft | microsoft_edge_for_android | >= 1.0.0 < 145.3800.99 | 145.3800.99 |
| microsoft | microsoft_edge_for_ios | >= 1.0.0.0 < 145.3800.99 | 145.3800.99 |
| microsoft | microsoft_excel_for_android | >= 16.0.0.0 < 16.0.19822.20038 | 16.0.19822.20038 |
| microsoft | microsoft_excel_for_ios | >= 1.0 < 2.106.26020617 | 2.106.26020617 |
| microsoft | microsoft_loop_for_ios | >= 2.0.0 < 2.106.26020617 | 2.106.26020617 |
| microsoft | microsoft_onenote | >= 1.0.0 < 2.106.26020617 | 2.106.26020617 |
| microsoft | microsoft_onenote_for_android | >= 16.0.1 < 16.0.19725.20142 | 16.0.19725.20142 |
| microsoft | microsoft_outlook_for_android | >= 1.0 < 5.2605 | 5.2605 |
| microsoft | microsoft_outlook_for_ios | >= 1.0.0 < 5.2605 | 5.2605 |
| microsoft | microsoft_outlook_for_mac | >= 1.0.0 < 5.2605 | 5.2605 |
| microsoft | microsoft_powerbi_for_android | >= 2.0.0 < 2.2.260210.21290750 | 2.2.260210.21290750 |
| microsoft | microsoft_powerbi_for_ios | >= 1.0.0 < 1.2.260302.2193910 | 1.2.260302.2193910 |
| microsoft | microsoft_powerpoint_for_android | >= 16.0.0.0 < 16.0.19822.20038 | 16.0.19822.20038 |
| microsoft | microsoft_powerpoint_for_ios | >= 1.0 < 2.106.26020617 | 2.106.26020617 |
| microsoft | microsoft_teams_for_android | >= 1.0.0 < 1.0.0.2026043102 | 1.0.0.2026043102 |
| microsoft | microsoft_teams_for_ios | >= 2.0.0 < 8.3.1 | 8.3.1 |
| microsoft | microsoft_word_for_android | >= 16.0.0.0 < 16.0.19822.20038 | 16.0.19822.20038 |
CVSS provenance
nvdv3.17.1HIGHCVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N
vendor_msrc7.1HIGH
Stop checking back — get the weekly exploitation signal.
Every Monday: what got weaponized or added to CISA KEV in the last seven days — each CVE cross-linked to its PoC, Nuclei template, and detection rule. Free, one email a week, unsubscribe in one click.
Microsoft
M365 Copilot Information Disclosure Vulnerability
vendor_msrc·2026-03-10·CVSS 7.1
CVE-2026-26133 [HIGH] CWE-77 M365 Copilot Information Disclosure Vulnerability
M365 Copilot Information Disclosure Vulnerability
Description: AI command injection in M365 Copilot allows an unauthorized attacker to disclose information over a network.
FAQ: According to the CVSS metric, the attack vector is network (AV:N) and user interaction is required (UI:R). Why does the CVE title indicate that this is information disclosure?
An attacker who successfully exploited this vulnerability could use malicious email to cause Copilot to present authoritative‑looking phishing messages that prompt the user to click links leading to data exfiltration or navigation to a malicious site.
FAQ: According to the CVSS metrics, successful exploitation of this vulnerability could lead to major loss of confidentiality (C:H), and some loss of integrity (I:L), but no loss of availabili
GHSA
GHSA-rjf5-cxrf-4rvw: AI command injection in M365 Copilot allows an unauthorized attacker to disclose information over a network
ghsa_unreviewed·2026-03-16
CVE-2026-26133 [HIGH] CWE-77 GHSA-rjf5-cxrf-4rvw: AI command injection in M365 Copilot allows an unauthorized attacker to disclose information over a network
AI command injection in M365 Copilot allows an unauthorized attacker to disclose information over a network.
No detection rules found.
No public exploits indexed.
2026-03-16
Published