How does human brain represent safety? How do we compute our level of safety through self-oriented and other-oriented calculation? How are our safety computation biased during integration of multiple self/other features? In this fMRI study, we examine how human brain represent internal and external information separately, as well as the metacognitive processes involved to compute one's safety level integrating internal and external information. We focus on the role of vmPFC in integrating self/other oriented information to represent safety, guiding human survival decisions under threat.