Ciphertext integrity is a notion which closely resembles message authentication codes (MACs) and is the cipher analogue of CMA-security for them.
The adversary Eve is given oracle access $\textit{Enc}_k$ and can query it with $q$ messages $m_1, m_2, ..., m_q$ in order to obtain their ciphertexts $c_1, c_2, ..., c_q$. Her goal is to produce a new valid ciphertext $c \notin \{c_1, c_2, ..., c_q\}$, i.e. a ciphertext such that $\textit{Dec}_k(c) \ne \textbf{error}$.
A cipher $(\textit{Enc}_k, \textit{Dec}_k)$ provides *ciphertext integrity (CI)*, if for all keys $k \in \mathcal{K}$, the probability that Mallory achieves her goal is negligible, i.e.
$$\Pr_{k \leftarrow_R \mathcal{K}}[\textit{Dec}_k(\textit{Eve}()) \ne \textbf{error}] \le \textit{negl(n)}$$
Similarly to MACs, Eve has access to a bunch of messages and their ciphertexts and she strives to produce a new valid ciphertext which does not cause the decryption function to error. A cipher has CI if she cannot succeed with significant probability.