CVE-2025-0740

8.6 HIGH

📋 TL;DR

An authenticated attacker can access other users' chat messages in EmbedAI by manipulating the CHAT_ID parameter in the load_messages endpoint. This affects all EmbedAI users running version 2.1 or earlier, potentially exposing sensitive conversations.

💻 Affected Systems

Products:
  • EmbedAI
Versions: 2.1 and below
Operating Systems: all
Default Config Vulnerable: ⚠️ Yes
Notes: All deployments with the vulnerable endpoint are affected regardless of configuration.

📦 What is this software?

⚠️ Risk & Real-World Impact

🔴

Worst Case

Mass data breach exposing all user conversations, including sensitive business communications or personal information, leading to regulatory fines and reputational damage.

🟠

Likely Case

Targeted access to specific users' conversations for espionage, blackmail, or competitive intelligence gathering.

🟢

If Mitigated

Unauthorized access attempts are logged and blocked, with no data exposure beyond intended recipients.

🌐 Internet-Facing: HIGH
🏢 Internal Only: MEDIUM

🎯 Exploit Status

Public PoC: ✅ No
Weaponized: UNKNOWN
Unauthenticated Exploit: ✅ No
Complexity: MEDIUM

Requires authenticated access but exploitation is straightforward via parameter manipulation.

🛠️ Fix & Mitigation

✅ Official Fix

Patch Version: EmbedAI version 2.2 or later

Vendor Advisory: https://www.incibe.es/en/incibe-cert/notices/aviso/multiple-vulnerabilities-embedai

Restart Required: No

Instructions:

1. Download EmbedAI version 2.2 or later from official sources. 2. Replace vulnerable files with patched versions. 3. Verify endpoint access controls are functioning.

🔧 Temporary Workarounds

Endpoint Access Restriction

all

Temporarily restrict access to the vulnerable endpoint using web server rules or WAF.

# Example nginx location block
location /embedai/chats/load_messages { deny all; }

🧯 If You Can't Patch

  • Implement strict session validation and user authorization checks before processing chat_id parameters.
  • Monitor and alert on unusual access patterns to the load_messages endpoint.

🔍 How to Verify

Check if Vulnerable:

Attempt to access another user's chat messages by modifying the chat_id parameter while authenticated as a different user.

Check Version:

Check EmbedAI version in admin panel or configuration files.

Verify Fix Applied:

Test that modifying chat_id parameters no longer returns unauthorized chat messages.

📡 Detection & Monitoring

Log Indicators:

  • Multiple failed authorization attempts on /embedai/chats/load_messages
  • Successful access to chat_id values outside user's normal range

Network Indicators:

  • Unusual pattern of requests to load_messages endpoint with varying chat_id parameters

SIEM Query:

source="web_logs" AND uri_path="/embedai/chats/load_messages" AND status=200 | stats count by user, chat_id

🔗 References

📤 Share & Export