Detection queries Not classified

Railway October 2026 incident - build escape to decrypted customer AI provider keys

Railway (Jacob Cooper, Angelo Saraceno, Sarah Bedell, Luca Casonato; report published Nov 10, 2026) says the attacker's first signups began on October 18, when dummy accounts were created. At 07:28 UTC on…

What happened

Railway (Jacob Cooper, Angelo Saraceno, Sarah Bedell, Luca Casonato; report published Nov 10, 2026) says the attacker's first signups began on October 18, when dummy accounts were created. At 07:28 UTC on October 20 a previously unknown bug in the multi-tenant build infrastructure gave a partial filesystem escape and a build-server credential. Railway's internal Temporal cluster trusted that credential. The attacker read stored histories of 353 builds between 13:50 and 15:30 UTC on October 20, including sealed build variables, and pulled a snapshot of Railway's internal monorepo. Railway says the source let the attacker run agent fanouts, find exploits, and move laterally. At 03:26 UTC on October 21 a misconfigured OIDC binding in a cloud environment exposed a short-lived cloud credential, which led to an SSH key and a long-lived credential on an orchestration machine. Between 04:42 and 05:14 UTC, and again between 06:36 and 07:15 UTC, the attacker queried the envelope-encrypted variables database with regexes for AI provider key names, focusing on customers with many variables, and decrypted them through KMS. At most 1.85 percent (about 100,000) of workspaces were affected. Linked-account tokens, BYOK agent keys and GitHub OAuth tokens, were exposed for 1,852 users. Railway reset those tokens. Unauthorized spend on the stolen keys was estimated at about 740,000 dollars. Starting at 09:25 UTC on October 21, Railway asked GitHub, GitLab, Anthropic, OpenAI, OpenRouter, SpaceXSI, Google, and Stripe to revoke keys. Railway reads the downstream use as credentials sold through AI-access relay services for distillation, seen in coding-agent traffic, agent-behaviour reviews used as training data, and checks of training tasks and their scoring. Railway links the incident to the wider Black October series (Fireworks, OpenCode, Warp, and others) without saying they are the same operator. The agent and ARTEX-like harness detail is attributed to Wiz. No attacker IPs or file hashes were published. ARTEX reference https://github.com/RuoJi6/ARTEX.

The attacker selected and decrypted customer AI provider keys. Railway, citing Wiz, also describes the intrusion itself as a contest against abliterated models and an ARTEX-like harness. That operator-side claim is vendor-asserted.

Read the original sourceAdditional source: Wiz: Black October for AI developer platformsAdditional source: BleepingComputer: Railway breach and fraudulent AI chargesAdditional source: The Hacker News: Railway stolen customer AI keys

How it happened

Select a stage to explore its defenses
Detect this stage
Detect this stage
Detect this stage
Detect this stage
Select a stage · Drag to explore · + / − to zoom · Dashed edges are loopsSelect a stage · Arrows show reported connectionsReported sequence
Read the flow as text & sources

A catalog reconstruction of the supplied source account. Unknown steps remain explicit; an edge describes the reported relationship between steps.

  1. 01 · Dummy account swarm

    Signups begin October 18, two days before the build escape.

    Hunt: Repeated credential attempts or a stolen session are followed by a successful login or resource access inconsistent with the owner.

    Shared infrastructure and legitimate travel create noise. IoT, SharePoint and cloud identity require their own collectors and field mappings.

    Source for step 1
  2. 02 · Break out of the build

    At 07:28 UTC on October 20 a build bug yields a credential that Temporal trusts. 353 build histories are read.

    Source for step 2
  3. 03 · OIDC misbinding to a long-lived credential

    At 03:26 UTC on October 21 a short-lived cloud credential exposes an SSH key and a long-lived credential on an orchestration machine.

    Hunt: A service key or role is used from an unfamiliar origin, gains broader access or causes inference usage outside its established baseline.

    Cloud control-plane logs alone may not include model prompts or inference calls. No endpoint policy is claimed to stop cloud-only key misuse.

    Source for step 3
  4. 04 · Decrypt AI provider keys

    Regex selection and KMS decrypt in two windows on October 21 expose variables for at most 1.85 percent of workspaces.

    Hunt: An unexpected process or tool accesses credential stores and is followed by token use or outbound transfer inconsistent with the owner.

    Defender DeviceFileEvents is not a general file-read audit trail. Collect suitable object-access, EDR or application logs; legitimate developer tools also read secrets.

    Source for step 4
  5. 05 · Stolen usage

    Keys show up in coding-agent traffic and training-data collection. Unauthorized spend is about 740,000 dollars.

    Hunt: Sensitive reads are followed by outbound transfer, a public artifact, or a tool call to a destination outside the approved workflow.

    An outbound connection or high byte count alone is not exfiltration. Encrypted traffic may hide content, and bulk legitimate exports are common.

    Source for step 5

Why it matters

Reported entry point: Previously unknown bug in Railway multi-tenant build infrastructure, partial filesystem escape, build-server credential. Reported impact: Keys show up in coding-agent traffic and training-data collection. Unauthorized spend is about 740,000 dollars; reported targets include Railway customers, at most 1.85 percent of workspaces; 1,852 users with linked BYOK agent keys or GitHub OAuth tokens; applicability to your setup has not been assessed, so verify affected versions, permissions, and configuration in the original report.

Read the original report

What you can do

Protection & prevention

Treat build credentials as production credentials. Do not let a build identity read Temporal histories or the variables database. Audit OIDC bindings. Alert on regex queries of variable names followed by KMS decrypt bursts. Prefer short-lived customer tokens and provider-side revocation.

Catalog defensive guidance based on the reported behavior. Confirm scope and current mitigations in the original source.

Review the original guidance Additional source 2 Additional source 3 Additional source 4
Investigate each phase

Detection queries

Draft query starters grouped by the attack stages where they apply.

Download stage pack

Start with the logs you collect. Each hunt identifies its required visibility, field mapping and limits. Queries are catalog-authored drafts using reusable behavior profiles; test with known activity and your normal baseline before alerting.

01Dummy account swarmCredential / session use outside baselineCustom schema · Draft

Signups begin October 18, two days before the build escape.

What to look for

Repeated credential attempts or a stolen session are followed by a successful login or resource access inconsistent with the owner.

Visibility: Defender-side logs where collected

Required logs

  • Identity / service authentication logs
  • Device, source-IP and session context
  • Resource access and token issuance audit

Coverage limit. Shared infrastructure and legitimate travel create noise. IoT, SharePoint and cloud identity require their own collectors and field mappings.

Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Credential / session use outside baseline
id: 7d73d919-e785-4d40-ab0b-f12949adfaf5
status: experimental
description: Repeated credential attempts or a stolen session are followed by a
  successful login or resource access inconsistent with the owner.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: application
  product: openattacks_audit
  definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
    Target and CorrelationId from the required source logs first.
detection:
  selection:
    EventType:
      - identity.login
      - identity.token.use
  condition: selection
KQL
Draft KQLDownload KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("identity.login", "identity.token.use")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated desc
SPL
Draft SPLDownload SPL
sourcetype="openattacks:audit"
EventType IN ("identity.login", "identity.token.use")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _time
Investigation steps
  1. Compare failures, successes and session reuse by account and source; inspect access to new resources.
  2. Check device registration, MFA and credential changes; do not equate travel anomalies with confirmed compromise.
03OIDC misbinding to a long-lived credentialUnexpected cloud / inference identity useCustom schema · Draft

At 03:26 UTC on October 21 a short-lived cloud credential exposes an SSH key and a long-lived credential on an orchestration machine.

What to look for

A service key or role is used from an unfamiliar origin, gains broader access or causes inference usage outside its established baseline.

Visibility: Defender-side logs where collected

Required logs

  • Cloud control-plane and identity audit
  • AI-service invocation / diagnostic logs when enabled
  • Usage, quotas, cost and key inventory

Coverage limit. Cloud control-plane logs alone may not include model prompts or inference calls. No endpoint policy is claimed to stop cloud-only key misuse.

Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Unexpected cloud / inference identity use
id: afac2e65-c838-485b-a26f-7f175d422f61
status: experimental
description: A service key or role is used from an unfamiliar origin, gains
  broader access or causes inference usage outside its established baseline.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: application
  product: openattacks_audit
  definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
    Target and CorrelationId from the required source logs first.
detection:
  selection:
    EventType:
      - cloud.credential.use
      - cloud.role.change
      - model.invoke
  condition: selection
KQL
Draft KQLDownload KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("cloud.credential.use", "cloud.role.change", "model.invoke")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated desc
SPL
Draft SPLDownload SPL
sourcetype="openattacks:audit"
EventType IN ("cloud.credential.use", "cloud.role.change", "model.invoke")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _time
Investigation steps
  1. Pivot on key ID / principal, source network and user agent, then inspect role changes, token minting and service pivots.
  2. Join inference usage and billing anomalies to the same principal; verify whether the behavior was authorized.
04Decrypt AI provider keysSensitive store access → credential useCustom schema · Draft

Regex selection and KMS decrypt in two windows on October 21 expose variables for at most 1.85 percent of workspaces.

What to look for

An unexpected process or tool accesses credential stores and is followed by token use or outbound transfer inconsistent with the owner.

Visibility: Defender-side logs where collected

Required logs

  • File-read / API access telemetry for relevant secret stores
  • Process identity and command lines
  • Identity, token and outbound-network audit logs

Coverage limit. Defender DeviceFileEvents is not a general file-read audit trail. Collect suitable object-access, EDR or application logs; legitimate developer tools also read secrets.

Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Sensitive store access → credential use
id: 46c46c62-2d83-4198-a8bf-d6637094527b
status: experimental
description: An unexpected process or tool accesses credential stores and is
  followed by token use or outbound transfer inconsistent with the owner.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: application
  product: openattacks_audit
  definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
    Target and CorrelationId from the required source logs first.
detection:
  selection:
    EventType:
      - secret.read
      - credential.store.read
  condition: selection
KQL
Draft KQLDownload KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("secret.read", "credential.store.read")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated desc
SPL
Draft SPLDownload SPL
sourcetype="openattacks:audit"
EventType IN ("secret.read", "credential.store.read")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _time
Investigation steps
  1. Identify the actor, process and exact store touched, then pivot to token use from new devices or infrastructure.
  2. Correlate reads with egress and rotate affected credentials after confirming exposure.
05Stolen usageSensitive data → unusual destinationCustom schema · Draft

Keys show up in coding-agent traffic and training-data collection. Unauthorized spend is about 740,000 dollars.

What to look for

Sensitive reads are followed by outbound transfer, a public artifact, or a tool call to a destination outside the approved workflow.

Visibility: Defender-side logs where collected

Required logs

  • Sensitive resource access / DLP records
  • Proxy, egress and cloud upload events
  • Agent / SaaS audit with correlation IDs

Coverage limit. An outbound connection or high byte count alone is not exfiltration. Encrypted traffic may hide content, and bulk legitimate exports are common.

Custom-schema candidates, not out-of-box detectors. Map your application / cloud / platform logs to EventType, Actor, Asset, Action, Target and CorrelationId. KQL uses a custom OpenAttacksAudit_CL table; SPL uses openattacks:audit. EventType values are proposed normalization names, not native vendor event names. Correlation and environment-specific alert logic remain to be built.

Sigma
Draft SigmaDownload Sigma
title: Hunt - Sensitive data → unusual destination
id: 04b60968-3179-4b1f-ae13-6f57c2446077
status: experimental
description: Sensitive reads are followed by outbound transfer, a public
  artifact, or a tool call to a destination outside the approved workflow.
author: MagicSword OpenAttacks catalog
date: 2026-10-23
falsepositives:
  - Authorized activity matching this broad hunt; baseline and investigate
    before alerting.
level: low
logsource:
  category: application
  product: openattacks_audit
  definition: Custom normalized audit schema; map EventType, Actor, Asset, Action,
    Target and CorrelationId from the required source logs first.
detection:
  selection:
    EventType:
      - data.export
      - agent.external.send
  condition: selection
KQL
Draft KQLDownload KQL
// SCHEMA TEMPLATE: create/map this custom table before use; not a built-in Sentinel table.
// OpenAttacksAudit_CL: TimeGenerated(datetime), EventType/Actor/Asset/Action/Target/CorrelationId(string).
// Selects candidates only. Apply the stage hypothesis and correlation steps before alerting.
OpenAttacksAudit_CL
| where TimeGenerated > ago(7d)
| where EventType in ("data.export", "agent.external.send")
| project TimeGenerated, EventType, Actor, Asset, Action, Target, CorrelationId
| order by TimeGenerated desc
SPL
Draft SPLDownload SPL
sourcetype="openattacks:audit"
EventType IN ("data.export", "agent.external.send")
| table _time EventType Actor Asset Action Target CorrelationId
| sort - _time
Investigation steps
  1. Join the resource read and outbound action by session, identity or process, preserving a bounded time window.
  2. Inspect destination ownership, payload classification and the user's intended action; baseline sanctioned sharing.
Technical detailsTooling, impact & MITRE ATLAS

Attack context

AI's role
AI credential theft
Additional role
AI as operator
Initial access
Previously unknown bug in Railway multi-tenant build infrastructure, partial filesystem escape, build-server credential

Tooling as reported

AI products, models and artifacts

  • Customer AI provider keys (Anthropic, OpenAI, OpenRouter, and others Railway asked to revoke)
  • ARTEX-like harness described by Wiz (https://github.com/RuoJi6/ARTEX)

Other tooling

  • Temporal client access with a build credential
  • Misconfigured OIDC binding
  • KMS decrypt of envelope-encrypted variables

Targets and reported impact

Targets / sectors
Railway customers, at most 1.85 percent of workspaces; 1,852 users with linked BYOK agent keys or GitHub OAuth tokens
Reported scale / exposure
About 100,000 workspaces (1.85 percent). 1,852 users' linked tokens. About 740,000 dollars of unauthorized model spend. 353 build histories read.
Target geography
Multiple regions

Reported attribution

Unnamed. Financial motivation (resale of AI key access). Railway links the case to Black October but does not say it is the same operator as Fireworks, OpenCode, or Warp.

Structured attribution

Actor
Unnamed
Country
Unknown
Motivation
financially-motivated
Attributed by
Railway; the agent-harness detail is from Wiz as cited by Railway
Attribution confidence
medium

Detection rules

  • Railway build-escape and KMS decrypt huntingGUIDANCE

    No indexed Sigma, Splunk, or Elastic rule matches Railway's private Temporal client, the build-isolation escape, or application-database KMS decrypts of customer variables. AWS rules for PutKeyPolicy, key deletion, or Kubernetes AssumeRoleWithWebIdentity describe different events and are not cited. On a platform like Railway, alert on a build identity reading job histories, on regex scans of variable names, and on decrypt bursts between new credentials and AI-provider key prefixes.

Activity timeline

First seen
2026-10-18
Last updated
2026-11-10

Updates

  • 2026-10-18 — Railway

    Attacker begins creating dummy Railway accounts

  • 2026-10-20 — Railway

    07:28 UTC build escape. 13:50-15:30 UTC the attacker reads 353 build histories, sealed variables included.

  • 2026-10-21 — Railway

    03:26 UTC OIDC misbinding. 04:42-05:14 and 06:36-07:15 UTC KMS decrypt of AI provider keys. 09:25 UTC revocation requests to GitHub, GitLab, Anthropic, OpenAI, OpenRouter, SpaceXSI, Google, and Stripe.

  • 2026-11-10 — Railway

    Full incident report. Unauthorized spend estimated at about 740,000 dollars. Wiz harness detail included.

Source defensive notes

Treat build credentials as production credentials. Do not let a build identity read Temporal histories or the variables database. Audit OIDC bindings. Alert on regex queries of variable names followed by KMS decrypt bursts. Prefer short-lived customer tokens and provider-side revocation.

MITRE ATLAS

AI-specific catalog annotations based on the reported behavior; these are not MITRE case-study assignments.

AML.T0021Establish Accounts

The attacker created a swarm of dummy Railway accounts starting October 18 Source

AML.T0055Unsecured Credentials

Regexes selected AI provider key names in the variables database before decryption Source

AML.T0012Valid Accounts

A build credential, then a cloud credential from a bad OIDC binding, were reused to move laterally Source

AML.T0034Cost Harvesting

Stolen keys ran up about 740,000 dollars in unauthorized model usage Source

AML.T0016Obtain Capabilities

Railway, citing Wiz, describes an ARTEX-like harness and abliterated models used during the intrusion Source

Sources & research credit

4 sources

Research by Jacob Cooper, Angelo Saraceno, Sarah Bedell, Luca Casonato · Railway

  1. Railway: Incident Report, October 21st Security Attack blog.railway.com
  2. Wiz: Black October for AI developer platforms www.wiz.io
  3. BleepingComputer: Railway breach and fraudulent AI charges www.bleepingcomputer.com
  4. The Hacker News: Railway stolen customer AI keys thehackernews.com

Times, counts, the 740,000 dollar estimate, the provider list, and the Wiz attribution of the harness are from Railway's Nov 10 incident report. No IPs or hashes were published. Link gtig-llmjacking-unc6993-2026-11 (Fireworks), keyhunter-llm-key-theft-2026-11 (Railway named only as a scanned PaaS), and wiz-ai-driven-cloud-2026-10.

Record & downloads
Source record name
Railway October 2026 incident - build escape to decrypted customer AI provider keys
Entry ID
railway-october-ai-key-theft-2026-10
Catalog review
November 12, 2026
Category
attacks-on-ai
AI involvement confidence
confirmed