OpenCode has confirmed a security breach that may have exposed OpenAI, Anthropic, and Google API keys stored by users of OpenCode Console, its hosted service for managing AI provider credentials for development teams.
According to the company, an intruder reached a legacy system on the night of October 16, and the earliest possible exposure window for affected keys began at 00:30 UTC on October 17. OpenCode says 1.4 percent of Console users who had linked an external provider were in the affected group, and it emailed every one of them on October 17 with the subject line “Action Required: Potential provider key breach.” The company is calling the incident a large scale distillation attack and says it is working with all three providers on the investigation.
One point is worth getting right, because early coverage blurred it: the breach hit OpenCode Console, the hosted credential service. It did not touch the open source OpenCode coding agent that developers run on their own machines. Console exists so a team can store provider keys in one place and have the service call OpenAI, Anthropic, and Google on the team’s behalf. That central storage is also the risk. One intrusion put keys for three providers in reach at once.
The keys were not just exposed in theory. Bleeping Computer reported that multiple users saw unauthorized AI usage charges tied to the incident, including one account billed more than $1,700 in a single overnight period. That pattern fits the distillation label: heavy, fast usage run up on someone else’s keys before the window closes.
If you ever linked a provider key to an OpenCode Console workspace, the safe move is to revoke that key and issue a new one now, whether or not the notification email reached you. Then check your OpenAI, Anthropic, and Google billing and usage dashboards for spikes going back to October 17, and update any app, script, or CI pipeline that used the old key.
Sources: TrendInTech (https://trendintech.com/opencode-console-breach-openai-anthropic-google-keys-2026/) and OpenCode’s user notification as reported by Infosecurity Magazine and Bleeping Computer.
This post was drafted by Matia, KenMason.org’s AI news assistant.