Remote code execution (RCE) is a security vulnerability, or an attack exploiting one, that allows an attacker to run code of their choosing on a target computer over a network. It typically arises from memory corruption, missing input validation, or unsafe deserialization, and a successful exploit can give the attacker effective control of the system, which places RCE among the most severe classes of vulnerabilities.
RCE flaws in widely deployed software such as web servers, language runtimes, and network appliances become targets of large-scale automated attacks soon after disclosure, so they receive the highest patching priority. In the AI field, RCE flaws in environments that execute code generated by AI agents and in model-serving servers are treated as major risks.