ShadowRay 2.0 Exploits Unpatched Ray Framework to Create GPU-Powered Mining Botnet

Oligo Security has reported ongoing attacks exploiting a two-year-old vulnerability in the open-source AI framework Ray. The attacks, dubbed ShadowRay 2.0, target clusters equipped with NVIDIA GPUs, transforming them into a self-replicating cryptocurrency mining botnet. This campaign is an evolution of a previous wave observed between September 2023 and March 2024, indicating a persistent and adaptive threat actor. The vulnerability in Ray allows attackers to gain control over the clusters, deploying mining software that leverages the powerful NVIDIA GPUs for efficient cryptocurrency mining. The focus on NVIDIA GPUs is strategic, as these devices offer high computational power, making them ideal for mining operations. The auto-replicating nature of the botnet enables rapid propagation across vulnerable clusters, exacerbating the impact. The implications of these attacks are significant. Organizations face resource theft, as their computational power is hijacked for mining, leading to potential financial losses. Additionally, the mining activity can disrupt legitimate operations by degrading cluster performance. The persistence of this campaign highlights the importance of timely patching and robust security measures. To mitigate the risk, organizations should immediately patch their Ray installations to the latest version. Implementing comprehensive monitoring and detection mechanisms is crucial to identify unusual activity, such as unexpected GPU usage or network traffic patterns indicative of mining. Strengthening access controls can also help prevent unauthorized access to the clusters. In conclusion, the ShadowRay 2.0 campaign underscores the critical need for proactive cybersecurity measures, including regular patching, monitoring, and access control enforcement. Cybersecurity professionals must remain vigilant and take immediate action to protect their AI infrastructure from such exploits.