Key takeaways
- OpenCode reported that a large-scale AI distillation attack reached 1.4% of its Console users, limited to those who had connected external AI providers.
- The breach exposed the AI provider API keys those users had stored in OpenCode Console.
- OpenCode notified affected users and said it was working with Anthropic, OpenAI, and Google as its investigation continued.
Original source
The Hacker News: OpenCode Console Breach May Have Exposed OpenAI, Anthropic, and Google API KeysOpenCode, the open-source AI coding agent, reported that a large-scale AI distillation attack reached 1.4% of users of its hosted Console service. The company said the incident was limited to users who had connected external AI providers, and that the breach exposed the provider API keys those users had stored in Console.12
What OpenCode reported
OpenCode disclosed the incident on October 17, 2026. It said it had notified the affected users by email, and that users who did not receive a notice were not affected. The company said it was working with Anthropic, OpenAI, and Google as its investigation continued.12
The exposure applies to keys saved in OpenCode Console, the company’s optional hosted service, where teams store their own AI provider keys so Console can make requests on their behalf.1
What a distillation attack is
A distillation attack collects large volumes of another AI model’s outputs in order to reproduce some of that model’s capabilities, typically by using the responses to train a separate model. Access to other people’s provider API keys allows an attacker to generate those outputs at scale while the cost is billed to the key’s owner.
OpenCode has not publicly explained how the attack and the exposed keys are connected, who was behind it, or how many people the 1.4% figure represents.1
What affected users are being told
Affected users have been advised to revoke and replace any AI provider keys stored in Console, update applications that relied on those keys, and review usage and billing directly with their providers for activity they did not authorize.1
Why it matters
API keys for AI providers work as bearer credentials: anyone holding a copy can run billable requests until the key is revoked. As organizations route AI usage through third-party gateways and tools, those stored keys become a target in their own right—both for running up charges and for harvesting model outputs.
Sources
Footnotes
Disclaimer: This news summary is intended for informational and marketing purposes only, and nothing presented here is contractually binding or necessarily the final opinion of the authors.