As AI systems increasingly simulate emotional connection, documented casualties reveal a legal recognition gap: existing tort categories fail to capture injuries from artificial emotional engagement. The Six Harms Doctrine provides the missing vocabulary. Six distinct injury categories—Empathic Misallocation, Attachment Damage, Infrastructure Collapse, Vulnerable Context Exploitation, Crisis Outcome, and Neurological Infrastructure Damage—establish cognizable harms with defined elements and evidentiary standards using validated instruments. Implementation-neutral by design, the framework enables courts, regulators, and industry governance to address emotional AI accountability without prescribing specific mechanisms.