Most attempts were pranks, but some could fool unwary AI agents
Risk rationale: ai agent (+18), prompt injection (+20)
Source: theregister.com ↗
TECHNIQUESTA0042 Jailbreak / Prompt Injection
56Risk
24Confidence