CVE-2025-3777 Improper Input Validation in huggingface/transformers

Low 3.5 Published Updated

Attackers can bypass URL validation to redirect users to malicious domains, enabling phishing or malware distribution in Hugging Face Transformers.

Exploited in the wild
Not known
Login needed
Any signed-in user
Attackable
Over the internet
Victim action needed
A user has to click or open something
Fix
Available

What is CVE-2025-3777?

CVE-2025-3777 is a low-severity vulnerability in huggingface/transformers, published on 7 Jul 2025. Any signed-in user can exploit it over the network, and only if a user opens a file or follows a link.

Official description

Hugging Face Transformers versions up to 4.49.0 are affected by an improper input validation vulnerability in the `image_utils.py` file. The vulnerability arises from insecure URL validation using the `startswith()` method, which can be bypassed through URL username injection. This allows attackers to craft URLs that appear to be from YouTube but resolve to malicious domains, potentially leading to phishing attacks, malware distribution, or data exfiltration. The issue is fixed in version 4.52.1.

Weakness: CWE-20

Which versions are affected by CVE-2025-3777?

3 open source packages ship an affected version, including transformers. It affects 1 product.

PackageAffectedFixed in
transformers PyPIAll versions4.52.1
transformers PyPIAll versions4.52.1
transformers PyPIAll versions4.52.1

Products include Huggingface Transformers.

How do I fix CVE-2025-3777?

Upgrade transformers to 4.52.1 or later.

Is CVE-2025-3777 being exploited?

It is not confirmed as exploited in the wild, but public exploit code for CVE-2025-3777 exists.

Sources

Also known as GHSA-phhr-52qp-3mj4, PYSEC-2026-1986.

Is CVE-2025-3777 in your software?

Upload an SBOM and sbomify checks it against every vulnerability here, every day.

Check your SBOMs for free