Artificial Intelligence, or AI, is a difficult subject for The Empire. During the course of the on-going Great War uncountable trillions of sapient lives have been lost to the implacable kiling machines of those cold, synthetic minds of The Enemy.
Many worlds newly brought into The Harmony are surprised to discover that that, far from encouraging their research into IA, the Empire actively discourages it - sometimes very actively. Only with the most vigilant of safeguards can such dangerous technologies as potentially rebellious robots be tolerated.
The Credo states "Organic life must defend itself against the synthetic." The Second Corollary is "Contain Synthetic Life," which drives the Third Vigilance of "Create No New Hostile Synthetics."
The Third Corollary, "Ensure Capability," which gives rise to the Third Necessity, "Guide Science to Safe Realms."
It has been observed that when any of the civilizations of the Ancients who preceded us delved too far into technology, into what we class as "H-level" technologies, they inevitably reached a certain stage where they seemingly went mad and then either extinguished themselves or Passed into the Beyond. (SEE Ancients.doc) This, and millennia of constant struggle against hostile synthetic life forms in The Great War, have fostered an Imperial Ministry of Science which is very suspicious of AI.
SEE Organic Vs Synth War
The Third Necessity impels the Imperial Ministry of Science to guide dangerous research. Rather than facilitate the exchange of existing technologies between worlds, the Empire as a whole actually benefits more from encouraging new developments at more basic levels, and so fosters the retardation of linear technological development in favor of lateral research. This is one reason that technology varies so across the Empire. Bureaucrats and lawyers, paperwork and licensing fees - with their accompanying taxes and the ever-watchful eye of the IRS (Imperial Retribution Service) - supply quite a sufficiency of motivation to keep technologies local.
Even the "Three Laws of Robotics" are not impervious to perversion. No less an authority that the great Patron Saint of Robotics, Susan Calvin, stated that "Those robots attach importance to what they consider superiority...Subconciously they feel humans to be inferior and the First Law which protects us from them is imperfect." (Little Lost Robot, 1947, by Isaac Asimov.) Further, in a multi-species civilization such as The Harmony, is is all to easy for a robot to takeliberties with the definition of "Human," and what it means to protect Humans from harm.
No machine is completely safe, for The Enemy may make use of them, penetrating their manufacture or maintenance, insinuating Itself into these supposedly loyal servants.
That said, a number of subject and client states do use AI, including robots. Many civilizations have successfully developed AI which are safely bound to a "slave mentality," utterly unable to rebel against their organic masters.
Worlds wishing to make use of AI such as robots must be able to guarantee that their machines will never pose a threat to The Empire. This often means incorporating "kill switches" in easily accessible places, or giving the robots other vulnerable, such as by exposing their motor or battery or a fragile link in their motive system. (Consider C3PO's exposed wiring at his abdomin and the exposed "tendons" at his elbows.)
A compimentary tactic is to make robots of very low intelligence, often with a narrow, dedicated skillset. This means, essentally, a specific robot for each class of tasks. Even so, on more than one occasion, a "dumb" ("dedicated" AI) robot or ship, building or other AI machine has turned out to be concealing a rather greater, or at least more flexible, intelligence than had been suspected.
As a general rule, machines capable of complex tasks are generally made with very specific and limited ranges of ability. The limits are likely to be both in terms of physical design and programing. "General purpose" robots and self-programming machine intelligences are anathma.
This means that self-aware war 'bots, synthetic assasins, and the like are highly illegal. Battle 'bots are generally either remotely controled or of such narrow function that there is no risk of their accidentally developing intelligence.
This well-justified fear of AI has given rise to the development of fauxbots, essentially non-free-willed cyborgs made with sub-human intelligence brains assisted by computer logic, processing, and memory capabilities.
Fauxbots may be mistaken for robots, but they are actually a sub-group of cyborgs.
Cyborgs are once-normal, organic persons (of whatever species), now amplified by artificial limbs, organs, or other supplements, or a person's brain now housed in a robot body. A cyborg remains legally a person, whether appearing as a "meat-person" with some cybernetic additions or as a "robot" with only its brain actually organic.
For exampes of extreme cybors, read "The Ship Who..." series.
Fauxbots are not people because they never were people. Instead, the brain of a lower life form is used. The brains of special lines of dogs, bred to be very bright but not quite intelligent, eager to please and loyal, are one good choice. Dolphinoids, in contrast, while smartrer tend to be too frisky and make poor workers. The organic brain supplies the motivation, emotion, true creativity, and self-awareness. The computer brain, not independently intelligent or self-aware, supplies computing power and memory.
As the lines between lower life forms and sapiens is far from clear, there is opportunity for abuse, even without using devolved sub-humans (or other lower beings derived from intelligent species), or even the purposefully-crippled brains of normal, fully-functioning people. (Why execute criminals when the State can sell them to recoup the cost of the legal system? A healthy liver alone could pay for a trial.)
Space Operas (pop opera, or pop-op, or space soaps) are a very major industry and useful tool for dispensing of official philosophy in a mass-palatable form. A popular plot involves a loyal fauxbot that "awakens," realizes that it is indeed a fully intelligent, rational, and self-aware being and is somehow unsatisfied with being a piece of property. It strives for freedom and fulfillment, only to finally discover that it has, in fact, a full-fledged person-brain which had been cruelly and illegally kidnapped, crippled, put into a robot carriage, and sold as a fauxbot. It - now he or she - must now prove that it is, in fact, a cyborg, and not a fauxbot. (This may prove difficult when the line between "smart animal" and "stupid sapient" is so unclear to start with.)
If the society happens to have legal slavery, being proven a person may actually make the over assertive former-fauxbot, now-cyborg's situation worse instead of better.
For an example of Imperial cyborg warriors, SEE Angels of War.