SGLang RCE Exploit Breaches NEAR AI, Steals Data From 18 Agents and About 100 Users

An attacker exploited a remote code execution vulnerability in the open-source SGLang inference engine, gaining access to one of NEAR AI's servers. The breach occurred last Sunday and allowed the attacker to run code on the server. As a result, the attacker may have been able to view requests being processed on the server, including those sent in confidential mode. Confidential mode processes requests within a trusted execution environment. The attacker stole data from 18 agents and around 100 users. The company disclosed the incident earlier this week.