AI Risk Wire

Cases › security breach › LR-2026-0872

NEAR AI security incident: SGLang remote code execution exposes user and legacy agent data

Who did it
Outside attacker
Who was hit
about 100 users' account information and data from 18 AI agents running in its legacy environment; two NEAR AI Cloud customers' API keys
Reported
10 Nov 2026
Severity
Moderate
  1. NEAR AI ran the open-source SGLang inference engine on its servers. It also operated legacy AI agents outside its confidential computing environment.
  2. Attackers exploited a remote code execution flaw in SGLang and ran code on one inference server. They then gained unauthorized access to internal infrastructure.
  3. Names and emails of about 100 users were downloaded, and data from 18 legacy agents was exposed. Two customers' API keys also leaked through logs.
  4. NEAR AI itself detected the unauthorized access.

Source: bloomingbit · 9 more

What was done

NEAR AI revoked sessions and admin credentials, replaced agent tokens, suspended Private Chat and patched SGLang. It notified affected users individually.

Withdrawn or disabled (Affected organisation) · Patched or fixed (Affected organisation) · Access revoked or keys rotated (Affected organisation) · Other stated action (Affected organisation) · Access revoked or keys rotated (Affected organisation) · Users or authorities notified (Developer or vendor)

Fix reported: “Revoked all active sessions and administrator credentials, replaced all agent tokens, suspended the Private Chat service, revoked API keys, and patched SGLang across all inference servers” · Short-lived, scoped credentials, Vulnerability scanning of AI components

Risks AI system attacked or breached (54 incidents)

No source states: AI company and model, money lost or records, tools it had, company using the AI, sector.

Evidence

Sources (10)

  1. NEAR AI Says Security Incident Exposed Data on 100 Users, Breached 18 Legacy AI Agents · bloomingbit · 2026-11-10
  2. NEAR AI Reports Data Breach Affecting Around 100 Users After AI Server Hack | ForkLog · ForkLog · 2026-11-10
  3. SGLang RCE Exploit Breaches NEAR AI, Steals Data From 18 Agents and About 100 Users · The Hacker News · 2026-11-10
  4. NEAR AI Halts Cloud Inference After Unauthorized Access to Internal Systems — Glitchwire · Glitchwire · 2026-11-10
  5. SGLang RCE exploit breaches NEAR AI, steals data from 18 agents and about 100 users · LavX News · 2026-11-10
  6. NEAR AI says an SGLang flaw let an attacker into its servers and expose about 100 users' account details · MLQ.ai · 2026-11-10
  7. Near AI Breach Exposes User Details and Data From 18 AI Agents · Yahoo Tech · 2026-11-10
  8. Across Protocol warns users to avoid app amid reports of drained wallets · TradingView · 2026-11-10
  9. Near AI Confirms Security Breach Affecting 100 Users and 18 Agents · Bitcoin.com News · 2026-11-10
  10. Near AI Confirms Security Breach Affecting 100 Users and 18 Agents · CryptoNews · 2026-11-11
Show every field, with its source and grade

What happened

Earliest source we hold
2026-11-10
D inferred by us from the described behaviour, reason shown
Who caused it
an outside attacker
D inferred by us from the described behaviour, reason shown
What was stopped
Private Chat service temporarily suspended; no duration stated
C one secondary source states it
What was done
Revoked all active sessions and administrator credentials, replaced all agent tokens, suspended the Private Chat service, revoked API keys, and patched SGLang across all inference servers
C one secondary source states it
What went wrong
security breach
D inferred by us from the described behaviour, reason shown

Impact

Severity
moderate
D inferred by us from the described behaviour, reason shown
Kind of harm
privacy, others' security
D inferred by us from the described behaviour, reason shown
Who was affected
about 100 users' account information and data from 18 AI agents running in its legacy environment; two NEAR AI Cloud customers' API keys
C one secondary source states it
Kinds of data exposed
personal information, credentials
D inferred by us from the described behaviour, reason shown
Why this severity
Other parties' data or systems were touched, and a source states how far it reached. Data of others was touched with a bounded scope: about 100 users and 18 AI agents.
D inferred by us from the described behaviour, reason shown

LR-2026-0872 · version 13 · published 2026-11-12 after automatic checks; not yet checked field by field by a person