Deniable encryption (DE) allows private communication over an insecure channel even under the coercion. That is, after an adversary forces communication participants to reveal their secret keys and randomness used during the communication, the message confidentiality can still be preserved. Since its introduction, a large body of studies have been made to improve the DE system in terms of efficiency, security definition and functionality, which is the focus of this paper. However, as far as we know, none of existing DE systems considers the abuse of deniability caused by malicious users, which is a crucial feature from the view of practical applications of DE. For instance, a malicious user (e.g., an employee who holds some sensitive documents and intends to sell these confidential contents for obtaining financial gain) can utilize a DE system to transmit the confidential content without the risk of being caught, which is extremely dangerous to the interests and security of his/her organization. In this paper, to mitigate this threat, we formally define the syntax and security notions of abuse-resistant deniable encryption, which restricts what users can deny. Then, we put forward a concrete construction of abuse-resistant DE scheme, and prove its security under the assumptions of indistinguishability obfuscation and one-way function. Compared with other related work, the proposed construction has advantages in terms of functionality and ciphertext rate simultaneously.