NEAR AI Confirms Breach Exposed User Data & Access to 18 AI Agents
NEAR AI has disclosed that its recent security breach exposed account information belonging to approximately 100 users, along with data from 18 legacy AI agents and API keys belonging to two cloud customers.
In a security incident report published November 10, the company said the attacker exploited a remote code execution vulnerability in SGLang, the open-source inference engine running on its servers. The flaw allowed the attacker to execute code on an inference server and reach part of NEAR AI’s internal infrastructure. The report fills in the picture NEAR AI first sketched in the early hours of November 10, when it confirmed unauthorized access, temporarily suspended cloud inference and later restored the service.
On November 8, attackers exploited SGLang flaw, accessing NEAR AI servers.
November 10 report disclosed breach, revealing 100 user accounts, 18 agents, two API keys.
NEAR AI patched vulnerability, revoked credentials, restored cloud service by November 10.
What Data Was Exposed in the Breach?
According to the incident report, the attacker reached an internal system tied to NEAR AI’s previously deprecated Private Chat service and downloaded account details for roughly 100 users. The information included email addresses, names and, for some accounts, subscription plan and balance details.
The investigation also confirmed access to 18 AI agents running on a legacy deployment that did not use NEAR AI’s Trusted Execution Environment (TEE). Data retrieved from those agents included chat histories, settings and operational logs. Separately, API keys belonging to two NEAR AI Cloud customers were exposed in logs; the company said both keys have been revoked.
NEAR AI said it found no indication that stored Private Chat conversations or uploaded files were accessed. It cautioned, however, that the attacker could have observed requests being processed on the compromised server on November 8, including requests running in confidential mode. The company did not specify how many requests may have been exposed.
SGLang Vulnerability Patched
NEAR AI said it responded by revoking active sessions and administrative credentials, replacing agent tokens and taking Private Chat offline. It deployed a fix for the SGLang vulnerability across all inference servers and completed the rotation of remaining credentials. The company said it had finished a forensic review, directly contacted users whose account details or AI agents were confirmed to have been accessed, and would make any regulatory notifications required.
The incident has highlighted a limitation in confidential computing. NEAR AI explained that its TEE infrastructure protects data from the system operator and host, but does not automatically guard against vulnerabilities in the software running inside the protected environment. In this case, the SGLang flaw let the attacker execute code within an inference server and potentially observe requests processed there, confidential ones included. NEAR AI said it had reported the vulnerability to the SGLang project and would now apply the same scrutiny to its inference software and supporting services as it does to the hardware-protected environment.
NEAR Protocol Remains Unaffected
NEAR AI reiterated that the incident did not affect NEAR Protocol or other NEAR services, and that NEAR AI Cloud was restored once the vulnerability was patched and affected credentials replaced.
While the report indicates no stored Private Chat conversations or uploaded files were accessed, the confirmed exposure of legacy agent histories, account details and customer API keys shows the breach reached further than the initial disclosure of unauthorized internal access suggested. The company has advised affected users to be wary of unsolicited messages referencing their accounts or balances, and said anyone whose account or agents were confirmed to have been accessed had been contacted directly.
Also Read: North Korean Hackers Laundered Binance Hack Funds via NEAR Meme Tokens: Elliptic
Disclaimer: The information researched and reported by The Crypto Times is for informational purposes only and is not a substitute for professional financial advice. Investing in crypto assets involves significant risk due to market volatility. Always Do Your Own Research (DYOR) and consult with a qualified Financial Advisor before making any investment decisions.


