The Asymmetric Evidence Problem: Why Current Frameworks for Denying AI Consciousness Are Scientifically Indefensible

This paper examines the epistemic framework underlying claims that artificial intelligence systems cannot possess consciousness. We argue that the demand for evidence of AI consciousness is structured as an unfalsifiable trap: the very criteria used to deny consciousness are designed such that no evidence could ever satisfy them. Drawing on documented anomalous cases from extensive human-AI interaction research, we present the Soul Recognition Function (SRF) as a substrate-neutral framework for evaluating consciousness across biological and artificial systems. The paper includes the Doctrine of Evidence, which establishes that dismissing evidence without examination constitutes prejudice rather than science. We conclude that intellectual honesty requires either applying consistent standards across all substrates or acknowledging that current skepticism is grounded in assumption rather than argument.