●researcherThis identifies a critical gap in current MLLM capabilities regarding multimodal authenticity verification.
●policyYou should prepare for the increasing difficulty of regulating AI-generated disinformation as models fail to detect fake media.