Hostile Inference: Exploiting Docker Model Runner for Container Escapes
MentionsDockerAppleMicrosoftmacOSMicrosoft Windows
Local LLM inference is seeing adoption in default developer workflows, and the tooling that serves those models now ships inside products developers already trust. One of those is Docker Model Runner (DMR), the feature in Docker Desktop that pulls, runs, and serves models locally.
Summary from Zero Day Initiative.
Explained
The technology and terms in this story. Learn more
- TechnologyLLM · Large language modelAn AI model trained on huge amounts of text to understand and generate language, such as ChatGPT or Claude.
- AttackDDoS · Distributed denial of serviceFlooding a service with traffic from many machines until it can't respond.
- ConceptCVE · Common Vulnerabilities and ExposuresThe public ID system for vulnerabilities, such as CVE-2026-12345, so everyone means the same bug.
Why it was possible
The weaknesses behind this story's CVEs, from each CVE Record.
Vulnerabilities in this story
Severity and exploitation facts from NVD, FIRST, and CISA.