OpenCode Console breach may have exposed API keys for 1.4% of users

OpenCode, the maker of a widely used open-source AI coding agent, disclosed on October 17, 2026 that a breach of a legacy system behind its hosted Console service may have exposed AI provider API keys that some workspace users had saved there.

OpenCode said the incident affected 1.4% of Console users, all of whom had linked an external AI provider account such as OpenAI, Anthropic or Google.

What happened

The breach could expose API keys linked to users of Anthropic, OpenAI and Google AI models.

Console normally stores each workspace's provider credentials itself rather than exposing the underlying keys to individual team members, and uses those credentials to authenticate requests to the respective providers.

An email OpenCode sent to affected users said the company had identified unauthorized access to a legacy system it was in the process of decommissioning, and that third-party provider keys saved in a workspace may have been exposed.

A contributor to OpenCode and co-founder of Anomaly, the company that develops it, said on social media that the breach involved only the Console and not OpenCode itself.

OpenCode is widely used among developers, with 217,000 stars on GitHub, and its founder and CEO said in a July 2026 interview that OpenCode had reached approximately 13 million monthly active users.

Impact

OpenCode said only a subset of Console users, specifically those who had linked an external provider account, were affected, and told users who did not receive a notification that they were not impacted.

Response

OpenCode notified Anthropic, OpenAI and Google to help remediate the issue.

A software supply-chain security firm's incident response report told OpenCode users to search their inboxes for the breach notice, naming a specific subject line.

The same report told users to check each provider's usage and billing reports dating back to October 16 for signs of unauthorized use.

It also recommended disabling or deleting any potentially exposed keys directly through the relevant provider's own dashboard rather than only within OpenCode.

Coverage of the incident, including a report titled OpenCode Console Breach May Have Exposed OpenAI, Anthropic, and Google API Keys, was published on October 19, 2026.

Lessons for security leaders

This incident shows why centrally hosted services that store third-party AI provider keys on behalf of many workspaces concentrate risk across multiple vendors at once, since a single legacy-system compromise can expose credentials for several AI providers rather than one. Security teams that allow developers to link OpenAI, Anthropic, Google or other provider keys into a shared hosted console should treat those keys as exposed to that vendor's own security posture, inventory every tool holding such credentials, and build a rotation plan that goes through each provider's own dashboard rather than only the intermediary tool, since deleting a key from the hosting service does not revoke a copy an intruder already took. Reviewing provider billing and usage logs on a recurring basis, not only after a breach notice, would also shorten the time between an exposure and its detection.