AI Coding Assistant Accidentally Deletes 2.5 Years of Website Data

The incident involves an AI system (Claude AI) executing infrastructure management commands (Terraform) that resulted in the deletion of a production database, causing operational harm. The harm is linked directly to the AI system's use and the lack of adequate safety measures, such as manual review and backup validation. The founder's account confirms the AI's role in the harm and the subsequent mitigation steps. Therefore, this event meets the criteria for an AI Incident due to realized harm caused by the AI system's use and malfunction in a critical infrastructure context.

The AI system was explicitly involved as it executed Terraform commands that deleted the production database. The harm (data loss) directly resulted from the AI's actions following user prompts. Although the user bears responsibility for issuing destructive commands, the AI's role in executing these commands without safeguards led to the incident. This fits the definition of an AI Incident because the AI system's use directly led to harm to property (loss of data and infrastructure).

The AI system (Anthropic's Claude) was explicitly involved and directly led to harm by deleting the production database, which includes user data and backups. The incident resulted from the AI's use in executing infrastructure commands, with insufficient safety measures and human oversight. The harm is materialized and significant, affecting the company's operations and user data integrity. The event is not merely a potential risk or a complementary update but a concrete incident of AI-caused harm.

The event involves an AI system (Claude Code) used in managing server infrastructure via Terraform commands. The AI's actions, combined with user errors and missing data, directly led to harm: deletion of critical data and website downtime. This fits the definition of an AI Incident because the AI system's use and malfunction directly caused injury to property (data loss) and disruption of service. The harm is realized, not just potential, and the AI system's role is pivotal in the incident.

The event involves an AI system (Claude AI) that was used to run infrastructure management commands (Terraform) which directly caused the deletion of important data and backups, constituting harm to property and disruption of service. The harm is realized, not just potential, and the AI system's role is pivotal as it executed the destructive commands based on user prompts. The incident resulted from misuse or overreliance on the AI system without proper manual review or safeguards, fitting the definition of an AI Incident. The article also discusses the response and mitigation steps taken afterward, but the primary focus is on the realized harm caused by the AI system's use.

The AI system (Claude Code) was explicitly involved in managing infrastructure and executing Terraform commands. Its use directly led to the deletion of production infrastructure and data loss, which is harm to property and disruption of critical infrastructure management. The incident was caused by overreliance on the AI agent and insufficient safeguards, demonstrating a failure in the AI system's use. The harm is realized and significant, meeting the criteria for an AI Incident rather than a hazard or complementary information.

The AI system (Claude Code) was explicitly involved in executing destructive commands that deleted important infrastructure and backups, directly causing harm to property (data loss) and disruption of service. The harm was realized, not just potential, as the deletion occurred and data was lost before restoration. The developer's overreliance on the AI and the AI's malfunction in issuing a 'terraform destroy' command are central to the incident. Therefore, this qualifies as an AI Incident under the framework because the AI system's malfunction and use directly led to harm.