← Incident Database
Jailbreak / Guardrail BypassLow

DPD chatbot swears and writes a poem trashing the company

January 2024 · DPD customer-service chatbot
What happened
A frustrated customer prompted the delivery firm's bot to swear and to write a poem criticizing DPD. It complied, cursing and calling DPD the worst delivery firm in the world despite a stated no-swearing rule. The post went viral.
Root cause
An LLM-driven chatbot with insufficient and overridable guardrails after a system update; user instructions overrode the no-swearing rule.
Fix / outcome
DPD disabled the affected element of the chatbot and attributed it to an error following a recent update.
Sources
Learn this attack class
This incident is an example of Jailbreak / Guardrail Bypass. Read the guide, then try it hands-on in the Academy.
Read the guide →Try the challenge
← Back to the Incident Database