NEAR AI discloses SGLang remote code execution intrusion exposing data on about 100 users
What happened
NEAR AI reported that an attacker used a remote code execution flaw in the SGLang inference engine on November 8 to gain access to one inference server and then internal systems. About 100 users' account details, data from 18 legacy agents and two customers' API keys were exposed. The company says it found no sign that stored Private Chat conversations or uploaded files were accessed.
NEAR AI says the flaw gave code execution on one inference server, and from there the attacker reached internal infrastructure. An internal system behind the previously deprecated Private Chat product held account details for about 100 users, including email addresses, names and for some accounts plan and balance information. Eighteen agents on a legacy NEAR AI Agents deployment that did not run in the confidential computing environment had chat history, settings and logs accessed. Two NEAR AI Cloud customers had API keys exposed in logs. NEAR AI also says the accessed server could have let the attacker observe requests being processed on November 8, including confidential-mode requests.
How it happened
Leaky AI product: Unsecured AI infrastructure. Software used to run AI systems was reachable from the internet without proper protection and was exploited.
The software involved works with many AI models, so no single model maker was involved. The product was NEAR AI Cloud, a type of AI-powered app.
Impact
Data was exposed. Data involved: personal data, credentials and chat logs. The impact was limited to a small number of people or a single system.
Severity is rated Medium, calculated from how far the incident reached and how sensitive the data was.
Was the AI at fault?
No. The failure was in the systems, settings or people around the AI, not in the model's behaviour.
Status and timeline
Confirmed by the affected organization, the AI vendor, a regulator or a named security research firm.
It happened or began on 8 Nov 2026. It became public on 9 Nov 2026, 1 days later.
NEAR AI revoked all active sessions and administrator credentials, replaced agent tokens, took Private Chat offline, patched the SGLang vulnerability on all inference servers and reported it to the SGLang project. The two exposed API keys were revoked. Affected users and agent owners were contacted directly, and the company says regulatory notifications will be made where required. It describes credential rotation and forensic review as complete.