The Trusted Platform Module (TPM) is a dedicated hardware chip designed to provide a higher level of security for computing platform.All TPM functionalities are implemented in TPM commands to achieve specific security goals.We attempt to analyze the security properties of these commands,especially the key management API.Our study utilizes applied pi calculus to formalize the commands and determine how their security properties affect TPM key management.The attacker is assumed to call TPM commands without bounds and without knowing the TPM root key,expecting to obtain or replace the user key.The analysis goal in our study is to guarantee the corresponding property of API execution and the integrity of API data.We analyze the security properties of TPM commands with a process reduction method,identify the key-handle hijack attack on a TPM newly created key,and propose reasonable solutions to solve the problem.Then,we conduct an experiment involving a key-handle attack,which successfully replaces a user key with an attackers key using malicious TPM software.This paper discloses the weakness of the rehtionship between the key handle and the key object.After the TPM software stack is compromised,the attacker can launch a keyhandle attack to obtain the user key and even break into the whole storage tree of user keys.