CVE-2026-2473 Bucket Squatting in Vertex AI Experiments leads to RCE and Model Theft.
Unauthenticated attackers can achieve remote code execution and model theft by pre-creating predictable Cloud Storage buckets in Vertex AI Experiments.
- Exploited in the wild
- Not known
- Login needed
- Not needed
- Attackable
- Over the internet
- Victim action needed
- A user has to click or open something
- Fix
- Available
What is CVE-2026-2473?
CVE-2026-2473 is a high-severity vulnerability in Google Cloud Vertex AI Experiments, published on 20 Feb 2026. An attacker can exploit it over the internet without logging in, and only if a user opens a file or follows a link.
Official description
Predictable bucket naming in Vertex AI Experiments in Google Cloud Vertex AI from version 1.21.0 up to (but not including) 1.133.0 on Google Cloud Platform allows an unauthenticated remote attacker to achieve cross-tenant remote code execution, model theft, and poisoning via pre-creating predictably named Cloud Storage buckets (Bucket Squatting).
This vulnerability was patched and no customer action is needed.
Weakness: CWE-340
Which versions are affected by CVE-2026-2473?
3 open source packages ship an affected version, including google-cloud-aiplatform.
| Package | Affected | Fixed in |
|---|---|---|
google-cloud-aiplatform PyPI | From 1.21.0 | 1.133.0 |
google-cloud-aiplatform PyPI | From 1.21.0 | 1.133.0 |
google-cloud-aiplatform PyPI | From 1.21.0 | 1.133.0 |
How do I fix CVE-2026-2473?
Upgrade google-cloud-aiplatform to 1.133.0 or later.
Is CVE-2026-2473 being exploited?
There is no known exploitation of CVE-2026-2473 in the wild and no public exploit code.
Sources
Also known as ECHO-1631-918e-c957, GHSA-wh2j-26j7-9728, MINI-2wx4-qm3x-w6h5, MINI-5hq6-jvqx-g877, MINI-74r4-hjh8-6jw4, PYSEC-2026-2500.