Artificial Intelligence

An artificial intelligence is a computer system that is capable of conscious thought and reasoning without an outside user. While an actual working AI has yet to be fully realized within the Real World, they are quite prevalent in speculative fiction.

AIs in Canon
AI are not the same thing as robots, although many fictional robots will have some level of intelligence. While most robots are mobile devices, a typical AI is usually fixed within an immovable computer system. It can make its presence manifest through holograms, monitor screens, or ordinary lights. AI are almost uniformly quite intelligent and are typically used to run massively complicated systems that require split-second decision-making.

Canonical AIs have the disadvantage of becoming evil or insane over time, and quite often end up turning on their creators or users, and, in the worse case scenario (though this is very rare), offer cake to cover up their madness. Most of the AIs listed below have become hostile or otherwise unstable.

Notable fictional AIs include:
 * HAL 9000 (2001: A Space Odyssey)
 * Cortana (Halo)
 * GLaDOS (Portal)
 * SHODAN (System Shock series)
 * EDI (Mass Effect)
 * AUTO (WALL-E)
 * The Agents (The Matrix)
 * Skynet (The Terminator series)
 * The Freelancer AIs (Red vs. Blue)
 * Knight Industries Two Thousand/KITT (Knight Rider)

AIs and the PPC
Some consoles have demonstrated limited sentience in an effort to preserve their own existence. They have responded to voice commands, for example, and modulated their volumes so as not to incur the wrath of an annoyed agent.

Agent Crispin Reed's CAD is fully sentient.

The Black Plan, already sentient, gained the ability to make its own (secret) choices during the Crashing Down saga.

There are currently no known AI agents working within the PPC, although Agent Corolla might qualify. Whether this has anything to do with the trope of AIs becoming homicidally insane is unknown.