Cases › security breach › LR-2026-0796
Stolen Anthropic and OpenAI API keys fuel model distillation black market
Preliminary Why: Only a republished copy: BigGo Finance · subject to change
Status: action reported, no fix
No source states: loss or records, developer and model, agent tool, could do, deployer, industry.
Severity moderate: test: data or systems of others touched with a bounded scope stated, the breach reached 1.4% of OpenCode Console users
Railway, a hosting platform whose systems store customers' environment variables, detected unusual activity at 1 a.m. on October 21. It told customers that an unauthorized third party may have accessed encrypted environment variables and, for some, obtained decrypted contents such as API keys and database credentials. Separately, the developer tool OpenCode said a breach reached 1.4% of its Console users, and its co-founder said attacks which used AI agents continued for at least 48 hours. Developers reported OpenAI and Anthropic keys consumed and accounts suspended. Railway invalidated all its API tokens and rotated storage-bucket credentials; its recovery was partial as of October 22. OpenCode rolled all keys and worked with the labs on refunds and account reinstatement. The link between the thefts and a black market for distillation data is the account of one commentator and is not confirmed by the affected companies.
What happened
What was done
Risks
Primary96 incidents
Rising: 24 vs 4.0 a week
Sorted by rule from the case's fields; not yet read from the sources.
Controls
Not yet read for controls. Controls that mitigate the primary risk:
ModelSafety training and refusal · Cyber and weapons safeguards · Abuse monitoring and account enforcement · Staged or restricted release · Dangerous-capability testing before release · Outside red-teaming · Content safety classifiers · Safeguards kept on in testing · Reasoning monitoring
Infrastructure / targetKnown flaws patched · Multi-factor sign-in and password policy · Network segmentation · Intrusion and endpoint detection · Backups and recovery · Phishing and fraud filters
Similar incidents
Evidence
No source states: what the AI was doing, what it reached or did, when it happened, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off.
Show the full account: 5 items from the primary sources
What happened, in order
- October 17 OpenCode publishes its account of a breach that reached 1.4% of its Console users.
- 5:30 p.m. on October 21 Railway deployments have been down for roughly 16 hours.
- 1 a.m. on October 21 Railway detects unusual activity in systems storing customer environment variables.
- morning of October 22 Railway's recovery is partial; the API is up but some features remain unavailable.
Not dated in the sources:
- Attacks, which used AI agents, continued for at least 48 hours according to OpenCode's co-founder.
Show every field, with its source and grade
What happened
Impact
Notes on gaps
Dollar figures for abuse and key value are estimates or reports and are not recorded as impact. Sources are press and a single creator's commentary; no primary incident report from Railway or OpenCode is included beyond quoted statements. Links and remediations for Anthropic and OpenAI controls are not recorded because no source reports them for this incident.
LR-2026-0796 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON