Going one step further, agents on an environment could actually have conflicting goals. Once we starting taking into account multiple agents competing for the same objectives, the field of game theory becomes important. Game Theory and reinforcement learning are the two fundamental fields of multi-agent reinforcement learning. When agents have opposing goals, there is probably no clear optimal solution and an equilibrium among the agents need to search for. For this, lots of game theory come to play.
Finally, we can think of worlds in which teams of agents compete against other teams for conflicting objectives. The RoboCup soccer is a well-known example of this type of environments. Make sure to check the recommended readings below, and work the RoboCup soccer provided by OpenAI Gym.