What happened

OpenCode, an open-source AI coding project, disclosed a security breach of its hosted service, OpenCode Console. It ties the incident to what it calls a large-scale distillation attack. It says about 1.4% of Console users were affected: those who had linked external AI providers such as Anthropic, OpenAI or Google. This is a breach of a hosted service, not a package or app release.

OpenCode emailed every affected user. Cyber Security News reports that third-party keys saved in affected workspaces may have been exposed, and that OpenCode described unauthorized access to a legacy system it was already preparing to shut down. BleepingComputer reports the earliest possible exposure was 00:30 UTC on 17 October and that attackers had targeted OpenCode's systems for at least 48 hours using AI agents.

BleepingComputer reports that users said they saw unauthorized charges from several hundred dollars up to $1,700, and that Google suspended some projects. These are user reports. BleepingComputer also says OpenCode's message mentions possible reimbursement of rogue charges.

The investigation is ongoing. OpenCode has not published a root cause, how long attackers had access, how many keys were involved, or whether exposed keys were used for distillation requests. It says it is working with Anthropic, OpenAI and Google. No indicators of compromise were published. Sources differ on the disclosure date, 17 or 18 October.

Timeline

  1. Cyber Security News reports OpenCode issued a statement on the breach and emailed affected users that morning.[3]

  2. Earliest possible exposure of users' keys, according to BleepingComputer's report of OpenCode's notice.[2]

  3. StepSecurity dates OpenCode's disclosure and user emails to this day, which conflicts with the 17 October date in other reports.[1]

Who is affected

  • OpenCode Console users who linked external AI providers (Anthropic, OpenAI or Google), about 1.4% of Console users.
  • Not everyone who has the OpenCode CLI installed; impact is scoped to Console accounts, per StepSecurity.
  • Users who did not get an email from OpenCode were not affected, according to OpenCode.

How to check

  1. Look for an email from OpenCode about the breach.
  2. Check whether you have an OpenCode Console account and which provider keys (Anthropic, OpenAI, Google) you stored or linked there.
  3. Review each provider's usage dashboard and your billing records for requests or charges you did not make, especially since 17 October.

What to do

  1. If you were notified, revoke and replace the Anthropic, OpenAI or Google keys you stored in OpenCode Console, using each provider's dashboard.
  2. Update every application that depends on the replaced keys.
  3. StepSecurity advises rotating any provider key connected to OpenCode or Console, including each provider listed in the local auth.json file.
  4. Report unauthorized charges to the provider and ask OpenCode about reimbursement.
  5. Watch OpenCode's updates for a root cause and any change to the scope.

Sources

Changelog

  1. First published.