User accounts disclosed to unauthenticated actors on the LAN

Description

Summary

The login page discloses all active user accounts to any unauthenticated browsing request originating on the Local Area Network.

Details

Starting the Home Assistant 2023.12 release, the login page returns all currently active user accounts to browsing requests from the Local Area Network. Tests showed that this occurs when:

  • The request is not authenticated and
  • The request originated locally, meaning on the Home Assistant host local subnet or any other private subnet (10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, fd00::/8, ::ffff:10.0.0.0/104, ::ffff:172.16.0.0/108, ::ffff:192.168.0.0/112)

The rationale behind this is to make the login more user-friendly (see release blog post) and an experience better aligned with other applications that have multiple user-profiles.

However, as a result, all accounts are displayed regardless of them having logged in or not and for any device that navigates to the server. This disclosure is mitigated by the fact that it only occurs for requests originating from a LAN address. But note that this applies to the local subnet where Home Assistant resides and to any private subnet that can reach it.

PoC

  1. Place a Home Assistant instance on a private subnet, i.e., 192.168.1.0/24.
  2. Create a few users, let's say, three.
  3. From any (or another) private subnet on the LAN, like 192.168.2.0/24, open an incognito browser window (to ensure that the browser has no cookies from Home Assistant and therefore is demonstrably unauthenticated) and navigate to the Home Assistant URL.
  4. The login page will display all three users, including their profile photo.

Impact

The following CVSS string could be shaped to describe the overall impact of this issue:
AV:A/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N

As seen, the Exploitability metrics are high, and the Impact metrics are low. This is fitting because the problem does not constitute a critical one, but at the same time, it is trivial to exploit. Still, since the mitigation can be so easily implemented in code to eliminate a typical case of information disclosure, it would certainly be worth pursuing.

Basic information

Type
reviewed
Severity
medium
Advisory on GitHub
Open advisory ↗
Repository advisory
Open repository advisory ↗
Source code
Browse source ↗
Published (advisory)
2023-12-15 23:19:28 UTC
Updated
2023-12-15 23:19:28 UTC
GitHub reviewed
2023-12-15 23:19:28 UTC
NVD published
2023-12-14

EPSS Score

Score Percentile
0.16% 37.10%

CVSS Scores

Base score Version Severity Vector
4.3 3.1
CVSS:3.1/AV:A/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N Click to expand
Attack vector (AV:A)
Attacker has to be nearby on the network—same office, same link, that vibe—not the whole wide internet.
Attack complexity (AC:L)
Once they can reach the bug, pulling it off is straightforward—no weird race conditions or rare setup.
Privileges required (PR:N)
No account or special rights needed—anonymous or random user is enough.
User interaction (UI:N)
Nobody has to click “OK” or open a trap file; it can work without a victim helping.
Scope (S:U)
Damage stays in the same “trust bubble” as the broken component—no big spill into unrelated systems.
Confidentiality (C:L)
Some sensitive info could get out, but not a total data dump.
Integrity (I:N)
Data isn’t meaningfully altered or forged.
Availability (A:N)
Service keeps running; no real outage angle.

Identifiers

CWEs

CWE id Name
CWE-200 Exposure of Sensitive Information to an Unauthorized Actor

Credits

  • r01k (reporter)

Affected packages (1)

Vulnerable version ranges and first patched releases as published by GitHub.

Ecosystem Package Vulnerable range First patched Vulnerable functions
pip homeassistant < 2023.12.3 2023.12.3

References

cvelogic Threat Intelligence