CVE-2026-73847 json
Emlog: Missing CSRF protection in AI Assistant execute_tool leads to full database compromise and admin account takeover
published 2026-08-14 2 mo ago modified 2026-09-16 54 d ago also EUVD-2026-58740
- cvss
- 6.8
- version 3.1, CNA
- cisa kev
- Not listed
- not in CISA's catalog
- epss
- 0.2%
- EPSS 0.2% chance of exploitation in 30 days, percentile 12%
- public exploits
- 0
- none known
Summary
Emlog is an open source website building system. In 2.6.26 and earlier, missing CSRF protection on the AI Assistant execute_tool action in admin/ai.php lets a remote unauthenticated attacker submit a forged cross-site request from an attacker-controlled page to a recently logged-in administrator. The authentication cookie set in include/lib/loginauth.php has no explicit SameSite attribute, enabling Chrome's temporary Lax+POST grace window. The query_database case passes attacker-controlled sql and confirm_code values to Ai::queryDatabase in include/service/ai.php; read queries need no confirmation, write queries accept the public confirm string, only the blog table is write-protected, and aliasing password as pwd_hash bypasses output redaction. A successful request can read every database table and write every table except blog, including changing the user table to take over an administrator account. No fixed version is available as of this review.
CISA SSVC
- Exploitation
- Proof of concept is public
- Automatable
- no
- Technical impact
- total
- Assessed
- 2 mo ago
Exploit prediction
0.22% chance of exploitation in the next 30 days, higher than 12% of all scored CVEs (FIRST EPSS, 2026-11-08).
Exploits and checks
- github.com https://github.com/emlog/emlog/security/advisories/GHSA-v6wr-4x55-7qp5 referenced PoC, tagged "exploit" by a source, unverified
CVSS scores
| assigned by | version | score | vector |
|---|---|---|---|
| CNA | 3.1 | 6.8 | CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:N |
Affected products
- emlog emlog
Classification
References
Sources
- CVE List CVE-2026-73847
- NVD CVE-2026-73847
Fencer Vulns keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.
CPEs and fixed versions use Anchore's correction of NVD's data: Added CPE configurations because not yet analyzed by NVD.