Critical vulnerabilities in LiteLLM expose risks in AI infrastructure
Security researchers at Check Point Research have identified two critical vulnerabilities in LiteLLM, a widely used open-source AI gateway. The flaws could allow attackers to gain administrative control, execute malicious code, and steal API keys from over 140 supported AI providers. With LiteLLM exceeding 90 million monthly downloads on GitHub, the vulnerabilities pose significant risks to organizations relying on the tool.
Editor, Lazyfounder
30 SEC SUMMARY
- Security researchers at Check Point Research discovered two critical vulnerabilities in LiteLLM, an open-source AI gateway used to connect applications to over 140 AI providers.
- The vulnerabilities (CVE-2026-35029 and CVE-2026-59823) could allow attackers to gain administrative control, execute malicious code, and steal AI provider API keys.
- LiteLLM has over 90 million monthly downloads on GitHub and is embedded in many platforms, amplifying the potential impact of these vulnerabilities.
- Users are urged to update to LiteLLM version 1.83.9 or later to mitigate risks, and organizations should rotate API keys if unauthorized access is suspected.
TABLE OF CONTENTS
KEY HIGHLIGHTS
- Check Point Research discovered two critical vulnerabilities (CVE-2026-35029 and CVE-2026-59823) in LiteLLM, an open-source AI gateway.
- CVE-2026-35029 allows low-privileged users to gain administrative control and execute malicious code, with a severity score of 8.7/10.
- CVE-2026-59823 enables server-side request forgery, potentially exposing AI provider API keys to theft.
- LiteLLM supports over 140 AI providers and has more than 90 million monthly downloads on GitHub.
- All versions earlier than 1.83.9 are vulnerable; users must upgrade immediately and rotate API keys if unauthorized access is suspected.
What LiteLLM does
LiteLLM is an open-source AI gateway that connects applications to large language models (LLMs) from providers such as OpenAI and Anthropic. It supports over 140 AI providers and is widely adopted, with more than 90 million monthly downloads and 61,000 stars on GitHub.
The tool is often embedded in platforms and internal tools used to build AI applications, increasing the risk of indirect exposure in environments where it is not deployed as a standalone solution.
Vulnerabilities discovered
Security researchers at Check Point Research identified two critical vulnerabilities in LiteLLM, tracked as CVE-2026-35029 and CVE-2026-59823.
CVE-2026-35029, rated 8.7/10 (high severity), allows low-privileged users to gain administrative control. Attackers could access admin functions, read server files, overwrite admin credentials, and execute malware.
CVE-2026-59823 enables server-side request forgery, allowing attackers to redirect API requests to servers under their control. This could lead to the theft of AI provider API keys, which may then be used to access other organizations’ AI accounts or incur charges on their behalf.
Recommended actions for users
LiteLLM has released a fix for both vulnerabilities in version 1.83.9. All users running earlier versions are urged to upgrade immediately to mitigate risks.
Organizations should review access logs for LiteLLM and rotate AI provider API keys if unauthorized access is suspected.
Security teams are advised to include AI gateways in routine security assessments, alongside web applications and cloud infrastructure, due to their growing role in enterprise AI deployments.
What this means
Lazyfounder analysis — our interpretation, not reported fact.
This incident underscores the growing security risks in AI infrastructure, particularly in open-source tools that act as gateways to multiple AI providers. For founders and operators, the vulnerabilities in LiteLLM serve as a reminder that AI-related components—especially those integrated into broader platforms—require the same rigorous security scrutiny as traditional enterprise software. The potential for indirect exposure (e.g., embedded use in internal tools) complicates risk assessment, making it critical to inventory all dependencies and enforce timely updates. The recommendation to rotate API keys also highlights the cascading risks of compromised credentials in AI ecosystems.
Key takeaways
- AI gateways and similar infrastructure tools are high-value targets for attackers due to their central role in connecting applications to multiple AI providers.
- Open-source AI tools, while widely adopted, can introduce systemic risks if vulnerabilities are left unpatched, especially when embedded in other platforms.
- Organizations must treat AI infrastructure as part of their core security perimeter, including regular assessments, patch management, and credential rotation.
- The widespread adoption of LiteLLM (90M+ downloads) demonstrates how quickly vulnerabilities can scale, emphasizing the need for proactive security measures in AI deployments.
FAQ
What is LiteLLM?
LiteLLM is an open-source AI gateway that connects applications to large language models (LLMs) from providers like OpenAI and Anthropic. It supports over 140 AI providers and is widely used due to its flexibility and ease of integration.
Which vulnerabilities were discovered in LiteLLM?
Two critical vulnerabilities were discovered: CVE-2026-35029, which allows low-privileged users to gain administrative control and execute malicious code, and CVE-2026-59823, which enables server-side request forgery and potential theft of AI provider API keys.
How can organizations protect themselves from these vulnerabilities?
Organizations should immediately update to LiteLLM version 1.83.9 or later, review access logs for unauthorized activity, and rotate AI provider API keys if unauthorized access is suspected.
Why are AI gateways like LiteLLM considered high-risk targets?
AI gateways act as central hubs connecting applications to multiple AI providers, making them attractive targets for attackers. A single vulnerability can expose API keys, administrative controls, and sensitive data across numerous AI services.
Related on Lazyfounder
Sources
- TechRadar · 2026-10-14
Security flaws found in top open source AI gateway — LiteLLM vulnerabilties could affect millions of users
This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.
About the author
Editor, Lazyfounder
Tarun Mottlia edits LazyFounders, covering Indian startups, funding rounds, AI and product launches. Every story on the site is AI-assisted and checked against its cited sources before publication.
More stories by Tarun MottliaGet the LazyFounder Brief
Startup, funding and AI news in a five-minute read. Join the early-access list.
