OpenCode Security Review: Console Breach and Distillation Attack Risks

OpenCode Security Review of the October 2026 Console Incident

OpenCode, developed by Anomaly Innovations, disclosed on October 17, 2026, that unauthorized access occurred to a legacy Console system it planned to retire. The company linked the event to a large-scale distillation attack and stated that third-party API keys stored by some users in hosted workspaces may have been exposed. Affected accounts represented about 1.4 percent of Console users who had linked external providers such as OpenAI, Anthropic, or Google.

The breach window began overnight between October 16 and 17, 2026, with the earliest reported exposure around 00:30 UTC on October 17. OpenCode notified impacted users by email the same morning and posted a public statement on X later that day. The open-source CLI, desktop, and local installations remained unaffected because they store keys locally rather than in the hosted Console.

Root cause details, including storage encryption practices, have not been released. OpenCode described the incident as ongoing through October 19 and confirmed continued attack attempts after the initial disclosure.

Provider Actions and Reported User Impact

Anthropic revoked affected Claude keys, while Google blocked exposed Gemini keys. OpenAI collaborated with OpenCode to restore access for impacted customers and issued refunds for unauthorized usage. OpenCode stated that refunds for charges incurred during the attack are in progress.

Self-reported losses ranged from several hundred dollars to approximately $1,700 per account, though OpenCode has not published aggregate figures or confirmed totals. The attack focused on harvesting model outputs for distillation, a tactic providers have increasingly countered through usage monitoring and key rotation requirements.

Users who did not receive an email from OpenCode were not considered impacted. Those who stored keys in the Console should revoke them directly through each provider dashboard and review recent usage activity.

Practical Steps After the OpenCode Breach

Revoke and replace any exposed keys at the provider level first. OpenAI, Anthropic, and Google each provide dashboards that list active keys and usage history, allowing quick identification of anomalies.

Enable spend limits or usage alerts on provider accounts to cap potential losses. Many providers allow daily or monthly budgets that trigger notifications or automatic halts when thresholds are reached.

Review workspace configurations in any hosted service and remove stored provider keys where possible. Store keys locally in encrypted files or environment variables instead of web dashboards when feasible.

Monitor official OpenCode channels for follow-up statements, as the company indicated additional updates would follow the initial disclosure. No full post-mortem report had been published as of October 20, 2026.