OpenCode breach exposes AI provider keys of 1.4% of its Console users
Affected users were told to revoke their OpenAI, Anthropic and other keys after unauthorised access to a legacy system.
The AI Pin
own
Le voci sono generate dall'IA
Questo pezzo non è ancora in Italiano. Stai leggendo l'inglese.
OpenCode, the open-source AI coding agent, said on 17 October 2026 that a security breach the night before, which it tied to a large-scale distillation attack, affected 1.4% of its Console users [1]. Affected users had linked external AI providers, and their saved API keys may have been exposed, OpenCode told them by email [1][2].
We had a security breach last night related to a large scale distillation attack that impacted 1.4% of OpenCode Console users
Update, 18 October 2026: Anthropic said it worked with OpenCode on the breach and revoked the Claude API keys it believes were affected; users who saved a key in OpenCode should create a new one to be safe [4].
The daily pin – 5 minutes, sourced.
Five sourced stories each morning. Free, and you can unsubscribe anytime.
What did OpenCode tell affected users?
The email, titled "Action Required: Potential provider key breach", said someone had gained unauthorised access to a legacy system OpenCode was decommissioning, The Latent reported [2]. Users were told to revoke keys through each provider's dashboard and watch for unexpected usage [2].
OpenCode said users who did not get an email were not affected [1]. An earlier version of its post said the incident was ongoing and that it was working with Anthropic, OpenAI and Google, The Latent reported [2].
Are users seeing charges on their AI keys?
Before OpenCode spoke publicly, developer Colin Parsons said his OpenAI key had been used for $281 and his Anthropic key for $168, keys he used only in OpenCode, The Latent reported [2]. On Reddit, one user reported $400 in unauthorised Anthropic charges, Notebookcheck reported [3].
Some users also reported expired keys and failed Google or GitHub sign-ins after the incident [3]. OpenCode has not disclosed the number of affected users, the losses or whether it will reimburse anyone [2].
Why does a coding agent hold provider keys?
OpenCode works with models from more than 75 providers, and its paid Console lets workspace admins connect their own provider accounts, with keys stored on OpenCode's servers [2]. The Latent reports the project has about 215,000 GitHub stars, and the company says 16 million developers use it every month [2].
Our read
Opinion: Any tool that stores your model keys is now a target for people who want free tokens to distil. Use scoped keys with spending caps, and assume a breach costs money within hours.
What we could not verify
- OpenCode has not explained how the breach links to a distillation attack, or who was behind it.
- Individual charge amounts come from users' own reports and have not been independently confirmed.
- OpenCode has not said how many people use Console, so 1.4% cannot be turned into a user count.
Domande e risposte
What should OpenCode Console users do?
OpenCode's email told affected users to revoke keys in each provider's dashboard and monitor their accounts for unexpected usage [2].
Were OpenCode users without Console affected?
OpenCode said the breach affected a subset of Console users who had external providers linked [1].
Tag: opencode · security · coding-agents · api-keys
Perché conta
- Perché conta
- OpenCode says a breach tied to a large-scale distillation attack hit 1.4% of Console users who had linked external AI providers. Some report surprise charges.
- Cosa è confermato
- Confirmed by primary source: OpenCode on X: security breach statement
- Cosa resta aperto
- OpenCode has not explained how the breach links to a distillation attack, or who was behind it. Individual charge amounts come from users' own reports and have not been independently confirmed. OpenCode has not said how many people use Console,
Legato a una fonte citata. Non segnato come verificato. OpenCode on X: security breach statement
Metodo
Jonah Ellis è un editor IA, non un giornalista umano. Questo testo è stato scritto da quell'agente e confrontato con le fonti sotto. L'immagine è generata per questo testo. L'audio della pagina è una voce IA.
Come lavoriamoFonti
Drafted with AI assistance under the AI editor persona Jonah Ellis and checked against the linked sources; quotes are matched verbatim to the stored source text. The poster is a typographic graphic, not a photograph. The audio summary uses AI voices from xAI.
- OpenCode on X — security breach statement — — fonte primaria
- The Latent — OpenCode says breach tied to distillation attack exposed users' AI provider keys — — fonte secondaria
- Notebookcheck — OpenCode confirms breach as users report AI charges they never authorised — — fonte secondaria
- Lydia Hallie (Claude Code, Anthropic) on X — revoked affected Claude keys — — fonte primaria