perceptions and respond or act in a rational environment, that is to say, correctly and tend to maximize an expected result.

In this context, the rationality is the characteristic that has an election of being correct, more specifically, of tender to maximize an awaited result. This concept of rationality is more general and for that reason more appropriate than intelligence (which suggests understanding) to describe the behavior of the intelligent agents. For this reason the consensus in calling them rational agents is greater.

An intelligent agent can be a physical or virtual organization. Although the term rational agent talks about to artificial agents in the field of the Artificial intelligence, also it can consider rational agents to the animal including the man.

The intelligent agents describe themselves schematically like an abstract functional system. Therefore, the intelligent agents sometimes are called Abstract Intelligent Agents (AIA) to distinguish them of their implementations of the real world like computer science systems, the biological systems, or of organizations.

Some definitions of intelligent agents make emphasis in their autonomy reason why they prefer the term independent intelligent agent. And others (in particular, Russell and Norvig (2003) consider conduct directed to objectives as the essence of the intelligent thing and prefer a term taken from the economy " Racional" agent;

In Sciences of the Computation the term intelligent agent can be used to talk about to a software agent that has something of intelligence, independent of if she is not a rational agent by definition of Russell and Norvig. For example, used independent programs for attendance of an operator or mining of data (sometimes denominated robots) also are called "intelligent agents".