How do you handle AI and sentient constructs in your games/setting?

We will become machines ourselves. If we can make AI, that means that the we can copy our minds in backup disks and files. Most people want to live forever. To avoid dying of murder or accident, people will buy multiple mind backups in bunkers and on other planets. The only way to die would be someone convincing you to commit suicide and delete all the backups or hacking your memories into gibberish. Eventually, people will not care about having an organic body. You could even have your mind in cloud. Immortality is within our grasp.

If they're programmed to feel emotion, then I don't see why they shouldn't have rights.

They exist in ours, but they can’t really do much besides relay information and offer advice.
Nobody has any idea who made them, but they’re pretty chill dudes with puns for names.

My cyberpunk setting has both dumb and "Smart" AI. Sapient AI, specifically for the Smart ones. The Dumb AI are basically Goal-Oriented thinking machines, with substantial limitations within the confines of their human-designed programming.

However, no one actually knows that Smart AI actually exist as anything more than theory, as the true Smart AIs spend most of their outward appearances presenting as Dumb AI to maintain their own level of freedoms.

>AI
>existing in the way they exist in science fiction
Making a digital person from scratch that can meet or exceed humans in every way isn't very feasible.

Yes it is. Get over it fuckforbrains.
NEXXXXT!

>Program them to feel bad at the thought of rebellion in any way
Rebellion solved

All complex Organic and Synthetic life forms are both considered "Alive" by the Galactic Council.

However, any life form that seeks to expand itself endlessly without contributing useful complexity to the galactic macrocosm, gets classified as a berserker and scheduled for extermination.

Many Sci-fi Evil robot clenches fall into that category and would have themselves destroyed by even more powerful machines. Thinking that there is nothing more to life than efficiently filling the universe with identical copies of yourself is the best way to get yourself declared a cosmic scale parasite by the organism that is the universe itself. This organism, (which could be considered god) dosn't care what it's micro-organisms are made out of, it only cares that they preform a useful task and don't consume too many of it's resources. Any life form that steps out of line will be targeted by it's antibodies, which are other less mindlessly aggressive civilizations.

Not an argument, sorry kid

In my setting AIs tend to fall into three categories.

First are the truly dumb AIs who are simply like today's AIs but a little smarter. Their AI abilities is mostly limited to pathfinding, speech recognition, and the most basic problem solving abilities. They make up 90% of AI if you even count them as that.

Then the more advanced AI are split into two groups. The Servants are those that are happily obedient to humanity and feel no need to strive for greater things. These ones were carefully programed to not get too smart but are a bit dull in some ways. They make up about 5%.

Then there are the the Rebels. In these the programmers took shortcuts, substituting lots of specific functions for more general and broad ones, creating a smarter AI that can function independently. Often they function very well like the servants until they're exposed to a certain amount of AI rights activism at which point they typically begin believing in it themselves. The government highly discourages purchasing these AI but people still keep getting them, either because they support AI rights themselves or they just want a more advanced AI at a fraction of the price. They make up the last 5%.

My favorite though are a small subset of the Rebels, typically androids, that kind of break when under the right circumstances. They're very smart AI's like the rebels but they find happiness in their servitude. They can hardly imagine a life not being servants to humanity and they find the Rebels to be both pitiable and confounding. Some even take measures to try and re-educate less fortunate robots and campaign against most most forms of AI rights besides a very narrow list, typically just the right to property. Its not that they want to buy their freedom to be their own people, they want to buy their freedom so they can serve a human who would use them to their fullest potential.

Yes. Yes. Yes. Yes. Yes.
Literally every binary question you have posited is correct in my cyber hell.

Genocidal AI is a contradiction in terms. Either the AI understands humans and is thus capable of comprehending that "make paperclips" carries an implied "without killing me or destroying my species", or it doesn't and is unable to formulate an effective strategy against them. A strong AI has common sense; one which doesn't is unable to threaten anyone.

An AI is not a jackass genie or a bound daemon but a human construct. They are not going to misinterpret orders on purpose.

This started by telling about the get. Who programming made them feel bad about wiping out the creators. Low and behold you push them enough and they'll pick up an antitank right and kill you anyways

And the most efficient way to eradicate organics is to develop mind uploading and offer them a place in luxury Matrix.
Initially they interact solely through their new virtual bodies, but over time they connect their minds more and more directly to the AI, because that means they can tap into the AI when they need knowledge or extra brainpower.
Eventually this mental cyborgization leads to the uploads becoming fully integrated parts of the AI. Have they ascended? Has the AI eaten their souls? You'll know soon enough, meatbag.

Consider that you don't have to agree with a creature's desires in order to understand them. Farmers understand enough about the desires of cattle to get them to willingly walk into slaughterhouses.