Tag: litellm
14 articles
Vulnerabilities in LiteLLM Gateway Expose AI Systems to Takeover
With two high-risk vulnerabilities discovered in LiteLLM, a popular open-source gateway tool, hackers can easily hijack AI systems, allowing them to wreak havoc and cause serious damage. Anyone with an account on a vulnerable deployment can take over the gateway and exploit it for malicious purposes.
PoeLLM Malware Expands Crypto Botnet Via 3,400 Servers
Meet PoeLLM, a sneaky malware that's rapidly expanding its crypto botnet by hijacking 3,400 servers, with its operators using a clever poem-based trick to hide its command-and-control address. This cunning tactic lets the malware derive new addresses by tweaking a few words in a public GitHub poem.
Hackers Expose 32 Zero-Day Flaws in Pwn2Own Ireland Contest
In just one day, top ethical hackers at Pwn2Own Ireland 2026 uncovered a staggering 32 zero-day vulnerabilities, raking in over $368,000 in prizes and showcasing the impressive skills of today's cybersecurity experts. This impressive haul highlights the vast attack surfaces of modern technology and the talent of the global ethical-research community.
LiteLLM Gateways Expose Cloud Credentials to Risk with Default Admin Key
Thousands of LiteLLM gateways are leaving cloud credentials vulnerable due to a shocking security flaw: nearly 300 online gateways accepted a default admin key, granting hackers unrestricted access to sensitive data and cloud accounts. This easily exploitable weakness puts countless organizations at risk of devastating breaches.
Malicious LiteLLM Releases Expose Over 2,100 Organizations to Credential Theft
Over 2,100 organizations are at risk of credential theft due to malicious LiteLLM releases that harvested sensitive data, including environment variables, SSH keys, and cloud credentials, and sent it to an attacker-controlled domain. These compromised packages were live on PyPI for about 40 minutes on March 24, leaving a trail of potential exposure.
LiteLLM Vulnerability Chain Enables Low-Privilege Server Takeover
A shocking vulnerability chain in LiteLLM has been discovered, allowing hackers to hijack servers with just a low-privilege account, and experts warn it's a critical threat with a near-perfect CVSS score of 9.9. By chaining three distinct bugs, attackers can escalate their access to full admin rights and run code on the server.
LiteLLM Flaw Exploited in Wild, Enables Unauthenticated RCE
A high-severity flaw in BerriAI's LiteLLM, known as CVE-2026-42271, has been actively exploited, allowing unauthenticated users to execute commands remotely. This critical vulnerability affects LiteLLM versions 1.74.2 to 1.83.7 and has been deemed a major security risk.
Data Breaches Surge, Exposing Sensitive Info at AI Startups, Agencies
Data breaches are surging, with AI startups and agencies exposed, as seen in the alarming theft of 10 petabytes from a Chinese supercomputer and 4 terabytes from AI startup Mercor due to a supply-chain vulnerability. These incidents highlight the hidden risks of connecting data to AI models, creating sensitive blind spots that leave large data sets vulnerable to compromise.
LiteLLM SQL Flaw Exploited 36 Hours After Disclosure
A critical SQL injection flaw, CVE-2026-42208, was exploited just 36 hours after its disclosure, putting vulnerable LiteLLM versions at risk of unauthorized database access. The bug, with a CVSS score of 9.3, allows unauthenticated callers to reach a vulnerable database query through the proxy's error-handling path.
Hackers Exploit LiteLLM SQL Flaw for Sensitive Data Access
Within just 36 hours of being publicly disclosed, a critical SQL injection flaw in LiteLLM, known as CVE-2026-42208, was actively exploited by hackers, allowing them to access sensitive data without authentication. This alarming vulnerability highlights the importance of swift patching, with LiteLLM version 1.83.7 now available to fix the issue.
Malicious AI Gateway Exposes Data Through Supply Chain Breach
A recent analysis of LiteLLM, a popular AI gateway, revealed a supply chain breach that embedded malicious code designed to steal sensitive data, highlighting the vulnerability of even the most trusted components. This breach turned a multifunctional gateway meant to enhance AI agents into a vector for data theft, putting countless users at risk.
LiteLLM Exploit Turns Dev Machines into Hacker Credential Hubs
Your developer's workstation is the secret Achilles' heel of your enterprise, unwittingly morphing into a credential hub where sensitive authentication material is created, tested, and reused - making it a prime target for hackers. A recent exploit, dubbed LiteLLM, has already shown how these machines can be turned into treasure troves for threat actors.
LiteLLM Supply-Chain Compromise Exposes Mercor Data
A single faulty AI dependency can become a backdoor for attackers - as seen in the recent LiteLLM supply-chain compromise that exposed sensitive data, source code, and internal credentials at Mercor. This alarming incident highlights the risks of relying on third-party dependencies and the importance of securing your supply chain.
Mercor Hit in Widespread LiteLLM Supply-Chain Attack
Thousands of companies, including AI hiring startup Mercor, have been hit by a widespread LiteLLM supply-chain attack, marking the first publicly disclosed downstream casualty of a software supply-chain intrusion. This incident raises a critical question: how can organizations trust their tech toolchains when the chain itself can be compromised?