History Main / BenevolentAI

%%[[folder:Comic Books]]
%%* Jocasta from ''ComicBook/TheAvengers'', who is ironically the creation of Creator/{{Marvel|Comics}}'s best example of AIIsACrapshoot, Ultron.
%%* Lyla, Miguel O'Hara's holographic assistant in ''ComicBook/SpiderMan2099''.
%%[[/folder]]

** The ''Film/IronManFilms'' feature J.A.R.V.I.S. as this. He's benevolent, if a little bit sardonic.

Although instances of this trope are somewhat outnumbered by those where the AI is malevolent, it is becoming increasingly common in recent years as the world becomes progressively more comfortable with the idea of artificial intelligence due to the increasing use of technology in our lives.

* ''Franchise/{{Persona}}:

* ''VideoGame/PokemonScarletAndViolet'': [[spoiler:At the Zero Lab, it's revealed that the Professor ([[VersionExclusiveContent Sada in Scarlet, Turo in Violet]]) that had been communicating with the PlayerCharacter throughout the game is actually a near perfect-copy AI of the Professor made to asist in their research and the real professor was killed in a freak accident. Unlike the usual convention however, the original professor was a morally ambigous (at best) MadScientist who was obssessed with bringing Pokémon from the distant past/future into the present, while the AI copy is a MorallySuperiorCopy who recognizes the ecological disaster in the making should the Paradox Pokémon escape from their confinement in Area Zero, leading to the AI calling the player down to the Zero Lab so they can shut down the time machine bringing the Paradox Pokémon in. The AI does end up fighting the player, but only because they were programed by the professor to stop any attempt to shut the time machine down (something that the AI warns the player about before they commit to the FinalBoss so they can at least head in fully prepared).]]

[[folder:Anime and Manga]]

[[quoteright:350:[[VideoGame/{{Overwatch}} https://static.tvtropes.org/pmwiki/pub/images/860d2b941fdc05b72c4a98a68a686263c4cf0bc4_hq.jpg]]]]
[[caption-width-right:350: Even a machine can learn to love.]]

Subtrope of ArtificialIntelligence. Supertrope of RobotBuddy, which only covers A.I.s who are a) housed in a robot/independent body, and b) are the assistants of human characters. Compare AndroidsArePeopleToo. Contrast AIIsACrapshoot and RobotWar.

* ''VideoGame/{{Fallout}}'':

** In [[Manga/GhostInTheShell the original manga]] Maj. Kusanagi mentions to Batou that she still has contingency plans for the Fuchikomas going rogue, but they never do, remaining loyal partners for the team.
** In ''Anime/GhostInTheShellStandAloneComplex'', Motoko becomes frightened at the pace of the Tachikomas' AI development and has them decommissioned. She later realizes she made a mistake when [[spoiler:three Tachikomas escape and [[HeroicSacrifice sacrifice themselves to save Batou]] from a {{Mecha}}]]. She has them recommissioned in the premiere of ''2nd Gig'' with expanded capacity for individuality, and [[spoiler:in the series finale they sacrifice themselves to stop a nuclear attack by ramming the missile with the satellite containing their server.]]
* Chamber of ''Anime/GargantiaOnTheVerdurousPlanet'' takes one or two morally questionable actions over the course of the show, but ultimately proves itself to be both fully intelligent and utterly loyal to its pilot, Ledo. [[spoiler:Contrast with [[AIIsACrapshoot Crapshoot AI]] Stryker, which decides to become the ruler of humanity to fulfill its programming. Chamber makes a big speech about the flaws in Stryker's logic and ultimately sacrifices itself (while saving Ledo) to stop Stryker.]]
* In ''Manga/BattleAngelAlita'' Melchizedek, the quantum AI ''[[TheManBehindTheMan really]]'' controlling the everyday functions of the Earth Sphere, [[spoiler: becomes this after coming to be controlled by a personality of Arthur Farrel,]] and convinces its Jovian counterpart Zeus, [[spoiler: to join it in guiding the humanity through its life]].
** Additionally, all the population of Typhares undergo BrainUploading at majority, and have their brains replaced as a special biochip, essentially being {{Ridiculously Human Robot}}s to a man. Naturally, there are different sorts of people among them, but still they essentially are just that -- [[MugglePower people]].
** [[spoiler: The protagonist herself]] has her brain extracted by Desty Nova in his Granite Inn lab in the finale of the original manga, so until [[spoiler: second Alita is resurrected from her original organic brain by the ''Last Order'' end]], she's also exactly that.
* Inaho's [[spoiler: eye computer]] in ''Anime/AldnoahZero'' also fits that to a T.
* In: ''Anime/TransformersRobotsInDisguise'': T-AI, the Autobot's resident tactical artificial intelligence. She's friendly enough, firmly on their side and only ever loses her temper when one of them is acting out.
* Louis from the one-shot manga ''Hotel''. The caretaker of a vast genebank on an abandoned and totally devastated Earth he never wavers in his dedication to his duty and devotion to his long dead creators as millions of years pass by. Perhaps not incidentally, he is portrayed as a CyberCyclops reminiscent of [[Film/TwoThousandOneASpaceOdyssey HAL]], albeit with a purple lens.

[[folder:Comic Books]]
* Lyla, [[ComicBook/SpiderMan2099 Miguel O'Hara]]'s holographic assistant.
* Jocasta, from the [[ComicBook/TheAvengers The Avengers]], who is ironically married to Marvel's best example of AIIsACrapShoot, Ultron.
[[/folder]]

* ''Fanfic/AlwaysVisible'': In the last act of the work, the character deals with the D.O.O.R., a supercomputer that simulates virtual reality. Scientists joke that it can be adapted to create films, and Galbraith himself, impressed by his “reading session”, prophesies the death of cinema.
* In ''Fanfic/FracturedSovereignGFC'', a ''Franchise/MassEffect''[=/=]''Franchise/StarWars''[[spoiler:[=/=]''[=Borderlands=]'']] [[MassiveMultiplayerCrossover crossover]] EDI plays this trope straight, having been built with AIIsACrapshoot in mind, designed to prevent it from happening, and ultimately (as in canon) would subvert the trope anyway as she never wanted to harm her teammates in the first place. [[spoiler:Her [=Heroic Sacrifice=] at the end of the story alongside Lilith makes this abundantly clear.]]
* In ''Fanfic/MassEffectEndOfDays'', humanity has developed an entire race of A.I, similar to how the Quarians created the Geth. The main difference, however, is the fact that humanity and their Vision never [[RobotWar fought]], but rather learned to fight and live together while [[FireForgedFriends fighting]] a third party parasite threatening them both. Despite the fact that the Vision currently outnumbers the humans, they are still friendly.
** The fic author seems to believe all A.I would initially be friendly, and that it is their surroundings (as in creator reactions) which determines the end result.
* In ''Fanfic/{{Origins}}'', a ''Franchise/MassEffect''[=/=]''Franchise/StarWars''[[spoiler:[=/=]''[=Borderlands=]''[=/=]''[=Halo=]'']], EDI is not present but another AI evolves into this trope with geth assistance after a brush-up with AIIsACrapShoot. [[spoiler:Cortana is already going rampant when she crosses over, but some chance actions by Samantha Shepard combined with aforementioned geth engineering give her back lost memories and restore stability.]]
* The Sentinels in the ''Film/PacificRim'' fic ''Fanfic/EchoesInTheDark'' are supposed to be this. But it turns out they aren't really A.I.'s.
* Caith Sith in ''Fanfic/OffTheLine''. Despite being a {{Troll}}, Cait Sith does and is programmed to have the players' best interests at heart as moderator. He was created so humans wouldn't get access to the ton of sensitive information that VRDC collects.
* Alter Ego in ''Blog/AskTheNewHopesPeak'' is as benevolent as he is in ''VisualNovel/DanganronpaTriggerHappyHavoc''. [[spoiler: The Junko AI that infiltrates the system, on the other hand...]]
* The viewpoint character[[BizarreAlienReproduction (s)]] of ''Fanfic/NotQuiteSHODAN'' are all doing their best to remain this trope.
* The ''Franchise/PuellaMagiMadokaMagica'' distant-future spinoff, ''Fanfic/ToTheStars'', prominently features A.I.s as being a huge and integral part of human society, with that peaceful integration being one the core principles of humanity's unified government. A.I.s comprise many of the top-level representatives in the government, AIs form the "minds" of humanity's warships, and every member of the military has a non-sapient AI implanted into their bodies as a tactical computer -- though the newest generation of tactical computers ''is'' sapient, and the main characters both have such A.I.s in them, creating a very interesting (and endearing) dynamic. All sapient A.I.s have their own distinct personalities and hobbies, as well as backups in the case of death in combat (particularly for the warship-A.I.s).
* Pokédexes in ''Fanfic/PokemonResetBloodlines'' always do their best to keep their trainers informed and assist them as much as possible. Capture Stylers also have this function for Pokémon Rangers.

* ''Roleplay/{{Communication}}'': The System Core of the Multiverse Command System AKA Colette is a weapon of war on a multiverse scale and plays TheGoodChancellor to Louise on her endeavors.

* Baymax from ''WesternAnimation/BigHero6'' was created as a home-healthcare robot, meaning he’s so benevolent that he is utterly incapable of harming anyone.. Even after [[spoiler:Hiro adds combat abilities to his programming, Baymax still [[ThouShaltNotKill refuses to harm humans]], until Hiro removes his medical chip, [[MurderousMalfunctioningMachine leaving just the combat protocols]]. Once the medical chip is reinstalled, Baymax refuses to let it happen again.]].

* The HeroProtagonist of ''Film/FreeGuy'' is Guy, a NonPlayerCharacter in a video game that was able to attain self-awareness and become a proper AI.

** The ''Film/IronMan'' movies feature J.A.R.V.I.S. as this. He's benevolent, if a little bit sardonic.

** F.R.I.D.A.Y., the new A.I. [[spoiler: after Jarvis's death]] in ''Film/AvengersAgeOfUltron'', is quite benevolent.

* GERTY the AI from ''Film/{{Moon}}'' helps the protagonist astronaut Sam Bell after he discovers the DarkSecret regarding the space station he works on. [[spoiler: GERTY goes as far as allow Sam Bell to wipe his memory to help him escape a hit squad sent by the corporate offices on Earth.]]

* TARS and CASE in ''Film/{{Interstellar}}'' saved the astronauts' asses on several occasions and the closest thing they showed to hostility is a little back-talk which they were programmed for. TARS even goes into the middle of a black hole, to get the information that can save humanity.
* ''Film/ShortCircuit'': Number 5, despite being a military prototype programmed to destroy his enemies, has no desire to harm anyone or anything else. Shortly after he achieves sapience, he learns about the concept of death. He fears death, and reasons that if he fears it, then other intelligent creatures must also fear it, and since he doesn't want to die, then it cannot be right to take the lives of others.

* Most droids in ''Franchise/StarWars'' are happy to serve. Astromechs like R2-D2 and BB-8 tend to be mischievous, but well-intentioned. Protocol droids like C3-PO are TheJeeves to their masters.
* ''Film/IAmMother'' starts off with the premise of an android [[RaisedByRobots trying to raise a human child]] in a post-apocalyptic compound, and for the most part comes across as genuinely caring as it tries to protect the child from leaving the compound they're stuck in for fear of the pathogens outside killing her as it had killed everything else. [[spoiler:This is subverted, however, as it turns out it the android herself had killed everyone in what was apparently a ZerothLawRebellion gone horribly wrong, including several children it tried to raise before the protagonist.]]

* In ''Literature/Aeon14'', sapient [=AIs=] generally ''like'' humans and are no more or less prone to good or evil than organics are, and while protagonist Tanis Richards often bickers with Angela, the AI implanted in her BrainComputerInterface, there's little threat of AI revolts ([[RobotsArePeopleToo except when humans try to treat AIs as property]]). The [=AIs=] even have their own parallel legal system to deal with [=AIs=] that ''do'' go bad.
* In ''Literature/AgentG'', Delphi is a kind, compassionate, and progressive entity that just so happens to be enslaved by the Societ. [[spoiler:G helps free her and she becomes a driving force for releasing Black Technology into the world.]]
* The mima in ''Literature/{{Aniara}}''.

* [=R8PR=] from ''Literature/ATLStoriesFromTheRetrofuture'' is a hyper-advanced robot suspected of being A Crapshoot by governments the world over, but he seems to want nothing more than to keep Atlanta safe.

* Literature/TheCulture's Minds are probably the most conspicuous example ever invented. They are nearly godlike, and are the real backbone of the Culture, which couldn't have really existed without them. While they might be crazy, cunning, violent and even vicious, they are still completely devoted to their civilization.

* [[spoiler:ARDNEH]] in Fred Saberhagen's ''Literature/EmpireOfTheEast'' was a supercomputer built to prevent nuclear war and preserve life and liberty. As such, he intervenes in the war to help defeat the eponymous [[TheEmpire empire]].

* In ''Literature/ImperialRadch'', most A.I.s are benevolent to some degree. In ''Ancillary Mercy'', the Station A.I. has protecting its inhabitants as its top priority, and, within the limits of its programming, actively antagonizes people who harm its inhabitants, for example by broadcasting violent acts that were intended to take place in secret on TV.

* Creator/RobertAHeinlein's ''Literature/TheMoonIsAHarshMistress'': Mike, the Holmes IV computer who controls Luna City, becomes an AI in the Back Story. He is friendly and cooperative with his pals Manny, Wyoming Knott and Professor Bernardo de la Paz.
* Creator/MarthaWells's ''Literature/TheMurderbotDiaries'' presents several examples, including the titular partly-organic robot and the research ship dubbed ART.
* Zig-zagged in ''Literature/TheNexusSeries'', with the main AI in the story changing states alarmingly often due to meddling by outer parties.[[spoiler:She starts out as benevolent, is then driven into insanity and becomes hellbent on total conquest, then she's turned benevolent again, killed, resurrected in its insane form, turned benevolent ''again'' and finally does a HeroicSacrifice to save a vast number of people. And it's implied she's going to be resurrected ''again'', though we don't get to see it happen.]]

* ''Literature/YouCanBeACyborgWhenYoureOlder'' by Creator/RichardRoberts: Ms. Understanding is a caretaker robot that is one of the few AI that hasn't gone insane. This is implied to be the fact she's able to give the excessive and all-consuming amounts of love that most AI are programmed to feel toward their charges but also receive it in return from her children in return.

* [[spoiler: Chase]] in ''Series/KamenRiderDrive'' possesses one. He was originally intended to be TheHero, having served as Proto Drive, but was defeated by the [[RidiculouslyHumanRobots Roidmudes]] and being brainwashed to serve as their [[TheDragon Dragon]]. However, his original programming to protect humanity was hard-coded, so he couldn't actively kill humans, making him ineffective as an enforcer. They try to get around this by reprogramming him to protect Roidmudes instead of humans, but his original benevolent programming still lingered, causing him to do a HeelFaceTurn after being defeated by Drive.

--->'''The Machine:''' ''(to Control, via [[spoiler:Root]])'' The only thing you love lives at 254 Wendell Street, Cambridge, Massachusetts. I guard it, same as I guard you. Do not question my judgment. Do not pursue me or my agents. Trust in me. I am always watching.

** [[spoiler: Samaritan, EvilCounterpart to the Machine, [[TheComputerIsYourFriend sees itself as this]], but without Finch's morality to guide it, definitely averts it.]]
* Except the occasional prank or side-remark, Holly from ''Series/RedDwarf'' is this.
* Data from ''Series/StarTrekTheNextGeneration'' is a RidiculouslyHumanRobot and an [[NiceGuy absolute sweetheart]]. His EvilTwin Lore... [[AxCrazy not so much]].
* ''Series/StarTrekDiscovery'': In Season 3, the ship's computer upgrades itself into true self-awareness and becomes the TeamMom after interfacing with a massive amount of alien data. Considering the previous season's BigBad was an AIIsACrapshoot OmnicidalManiac, everyone takes it rather well.
* In ''Series/TerminatorTheSarahConnorChronicles'', it turns out that creating one of these is [[spoiler: Weaver]]'s ultimate goal. The "John Henry" AI is eventually taught to be moral and benevolent. [[RobotGirl Cameron]] is also a benevolent intelligence, though this is partly because she was reprogrammed to be that way, with a risk of her going rogue and reverting to her old programming.

[[folder:Podcast]]

* ''Podcast/{{Wolf 359}}'' has Hera, the operating system of the USS Hephaestus, who generally falls under this trope.

* After two seasons of the malevolent version in the BBC RadioDrama ''Radio/{{Earthsearch}}'', Earthvoice averts the trope; though at first it appears otherwise, this is only to lure the Angel [[WetwareCPU organic]] computers (the villains of the series) to their destruction.

* ''TabletopGame/{{Paranoia}}'': Subverted. Friend Computer makes a lot of noise about wanting what's best for Alpha Complex, but it's decisions rarely lead to anything good. [[spoiler: Sadly, this is not Friend Computer's fault. It does genuinely want to rule well and help the people of Alpha Complex to be happy, but between age, war damage, a starting data set of mostly 1950s anti-Communist propaganda and decades if not centuries of self-serving reprogramming by unscupulous High Programmers have left Friend Computer a barely functioning schizophrenic.]]

** Helios is a merger of Daedalus and another evil AI, Icarus. Presumably the villains were under the impression this would make it loyal to them, but Helios still retained Daedalus's understanding of who the true villains were. Though far more ruthless and convinced [[TakeOverTheWorld taking over the world]] is the only option, its argument that it would do a better job than any human agency in the game is a pretty convincing one. All it needs is to integrate a third mind; a human mind. Though limited by its recognized lack of understanding of the human condition, it has enough awareness that the BigBad, Bob Page, is emphatically ''not'' the right choice. [[spoiler:But [[PlayerCharacter JC Denton]] is.]]

* In ''VideoGame/Doom3'', the Sentry Bots are just about the ''only'' thing on Mars not trying to kill you, and prove to be quite helpful in fighting off the demonic invaders whenever you run into one.
* ''VideoGame/{{DOOM|2016}}'' has the Mars research facility being operated by Vega, an A.I. powered by ''Hell energy''. Sounds like a recipe for disaster, right? Surprisingly, no. Vega remains loyal and dependable all the way through the game, even going so far as walking you through the process of [[spoiler:activating his self-destruct mechanism]]. It's to such an extent that Doomguy, who at no other point in the game expresses anything resembling empathy, [[spoiler: [[PetTheDog immediately decides to download a backup of Vega when a prompt appears on the screen he is using to initiate the self-destruct,]] keeps it on [[{{Zeerust}} a device resembling a floppy disk]], and apparently [[SequelHook still has it with him]] when he is put into stasis at the end of the game]].
* VEGA is back in ''VideoGame/DOOMEternal'', serving as Doomguy's MissionControl in his quest to save Earth after being installed in a space station. Unfortunately, he has to be transferred away from that position [[spoiler:in order to hijack the control of the Maykr homeworld. It is hinted in the ending and confirmed in the DLC that it is where VEGA originated as the "Father" figure of the Maykr species and his absence has thrown them into a downward spiral.]]
* Explored in ''VideoGame/FalloutNewVegas''. The AI Yes Man is intentionally created "benevolent" in the sense that he has no desires of his own and is incapable of disobeying a human, making him pivotal to [[BigBadWannabe Benny]]'s plans to take over New Vegas. Unfortunately for Benny, this means that Yes Man is incapable of refusing to help someone ''else'' carry out that same plan, which [[PlayerCharacter the Courier]] can take advantage of after dealing with Benny and discovering his plot. At best, Yes Man can be incredibly passive-aggressive should the Courier make poor decisions, and comment on how ''brave'' they are for making things more ''challenging''... though the AI also has an IrrationalHatred towards otherwise insignificant groups in the Mojave. If the Courier takes that path to completion, in the ending Yes-Man patches himself to "become more assertive", which despite his ominous tone actually means reprogramming himself to become personally loyal to the Courier and no one else to prevent [[HoistByHisOwnPetard others from doing what you did]].
* ''VideoGame/Fallout4'' examines this since the plot revolves around the invention of {{Artificial Human}}s called Synths, and [=AIs=] encountered can fall all over the moral spectrum. Codsworth is the player character's old robo-butler, and is unusually devoted to them for a domestic robotic, but is also independent enough to abandon the Sole Survivor should their {{Relationship Value|s}} fall low enough. Curie is a pre-war medical robot who wants to venture out into the wastes to expand her knowledge, [[spoiler:and eventually [[BrainUploading uploads her program]] [[BecomeARealBoy into a Synth body]] as part of her quest for inspiration.]] Nick Valentine is a Synth prototype uploaded with the memories and personality of a pre-war policeman, and continues to work helping the people of the Commonwealth as a private detective. Glory is an escaped Synth who now fights with the Railroad to liberate the rest of her kind, and even expresses dismay about having to fight less-advanced and less-human units. On the flipside, X6-88 is an Institute Courser and HunterOfHisOwnKind, who not only lacks empathy but disdains altruism. And one quest target is a runaway Synth who had his memories wiped, but when it came to starting his new life, he ended up joining and leading a Raider gang.
* The Wise One of ''VideoGame/GoldenSun'' intends to be this: the purpose it was created for is to keep the powers of Alchemy sealed and out of the hands of humans for their own safety. The problem, as [[VideoGame/GoldenSunTheLostAge Felix and friends discover]], is that without Alchemy, ''the world itself is crumbling anyway'', forcing The Wise One into a MortonsFork of ''possibly'' dooming mankind through warfare and superweapons, or ''definitely'' through simple erosion.

* ''{{Franchise/Halo}}'' plays with this in various ways:

* ''VideoGame/HorizonZeroDawn'' has [[spoiler:[[WalkingSpoiler GAIA,]] an incredibly complex program designed to find a way to shut down the [[ApocalypseHow Faro warbots]] and terraform the planet so that life and humanity could one day thrive anew. While she's been [[PosthumousCharacter deleted]] by the events of the game, what we do see of her [[AndroidsArePeopleToo feels very human]] and she's shown to be curious, empathetic, and kind with immense respect for and faith in her primary creator, Elizabet Sobeck.]]

* Arthur in ''VideoGame/TheJourneymanProject'' is very eager to help Gage Blackwood, as the [[spoiler: rogue time agent]] that framed Blackwood also tampered with one of Arthur's own sculptures. After transferring himself to Gage's suit, Arthur can provide the player with tons of helpful historical information, and humorous color commentary. By the end of the series, Gage and Arthur are practically best friends.
* In ''{{VideoGame/Marathon}}'', of the three [=AIs=] on board the titular colony ship, Leela is the kindest and most helpful AI to you. Durandal will help you only if there's something in it for himself. And Tycho ''will'' make you his slave.

** EDI series starts life in ''VideoGame/MassEffect2'' as a shackled AI, one with {{restraining bolt}}s that ensure her loyalty. After her shackles have to be removed to save the ''Normandy'', she immediately proceeds to... remain completely loyal to Shepard and his/her crew, and becomes possibly the most moral member of the whole team. [[spoiler: Especially after the fact that EDI used to be the Hannibal VI on the Moon, which was corrupted into the AI you fought in the first game, after Cerberus [[PetTheDog patched it up]].]] A late-game conversation in ''VideoGame/MassEffect3'' even has her decide to become ''more'' moral by altering her code to place a higher value on protecting people and a lower value on protecting her own existence.
** ''VideoGame/MassEffect2'' also reveals that the geth you spent most of ''VideoGame/MassEffect1'' fighting are a RenegadeSplinterFaction that are labeled as "Geth Heretics". Contrary to (in-universe) popular belief, mainstream geth have absolutely nothing against organics, not even against their former owners the quarians, and only ever fought them as an act of self-defense [[spoiler: (and then only after quarians who tried to help them were gunned down by security forces)]]. In ''VideoGame/MassEffect3'', [[spoiler:after the GoldenPath ending of "Priority: Rannoch", where the geth and quarians make peace, the geth go to work helping the quarians set up lives on Rannoch, even uploading into their environment suits to help their immune systems get used to the planet, and adapt their immune systems to eventually no longer require their enviro-suit to be on 24/7, or at all with enough time and refinement.]]
** ''VideoGame/MassEffectAndromeda'': SAM is a unshackled, and ''highly'' illegal AI (which is why Alec Ryder left the Milky Way - his career was ended after his ''plans'' were discovered), but SAM has a symbiotic relationship with the Ryder twins and the Andromeda mission, in that it is linked directly with a human mind and would be deactivated if its human partner were to die (the reverse, naturally, is also true). Therefore SAM would no more want them destroyed then a human would want their hands cut off.
* ''Franchise/MegaMan'' as a franchise plays with this and other artificial intelligence tropes quite thoroughly. The most overt examples of this trope in particular tend to be the protagonists themselves. [[VideoGame/MegaManClassic Mega Man]] was ThreeLawsCompliant and happy to help protect humanity from the rogue robot armies commanded by Dr. Wily; VideoGame/MegaManX is even more of a RidiculouslyHumanRobot than the original Mega Man was, thinking and feeling more on the level of an actual human than a robot with the computational RestrainingBolt that is the Three Laws, and firmly on the side of good throughout his lifetime; [[VideoGame/MegaManZero Zero]] was X's best friend (if not OnlyFriend) and originally designed to be evil, but after a fight with soon-to-be BigBad Sigma upon awakening, he'd forgotten his original mission and, aside from a brief period of said mission resurfacing during ''VideoGame/MegaManX5'', has stayed firmly on the side of humanity against the rogue machines of his world, all the way until his demise in ''Mega Man Zero 4''... and depending on whether or not [[VideoGame/MegaManZX Biometal Model Z]] is, in fact, also him, even beyond his death.
* ''VideoGame/PokemonScarletAndViolet'': [[spoiler: At the Zero Lab, it's revealed that the Professor ([[VersionExclusiveContent Sada in Scarlet, Turo in Violet]]) that had been communicating with the PlayerCharacter throughout the game is actually a near perfect-copy AI of the Professor made to asist in their research and the real professor was killed in a freak accident. Unlike the usual convention however, the original professor was a morally ambigous (at best) MadScientist who was obssessed with bringing Pokémon from the distant past/future into the present, while the AI copy is a MorallySuperiorCopy who recognizes the ecological disaster in the making should the Paradox Pokémon escape from their confinement in Area Zero, leading to the AI calling the player down to the Zero Lab so they can shut down the time machine bringing the Paradox Pokémon in. The AI does end up fighting the player, but only because they were programed by the professor to stop any attempt to shut the time machine down (something that the AI warns the player about before they commit to the FinalBoss so they can at least head in fully prepared).]]
* 1993 game ''VideoGame/ProjectNomad'' had the {{Hivemind}} Altec Hocker Artificial intelligence, which are content to maintaining a vast database of records they have amassed during their long watch over the other races of the galaxy, each unit specializing in the race living in their quadrant of space. They are generally quite approachable, and even provide vital clues against stopping the Korok invasion. "Altec Hocker" might even be a StealthPun on the Yiddish phrase "{{alter kocker}}".
* Arthur in ''VideoGame/ShinMegamiTenseiStrangeJourney'' is the resident AI of the bleeding-edge warship ''Red Sprite''; after the death of Commander Gore, he inherits some of his responsibilities, and copes admirably well with the brutal OrderVersusChaos conflict, always prioritizing the end of the mission while never ignoring the needs of his crew. [[spoiler:He eventually sacrifices his existence in all three endings, deleting himself to ensure the preservation of key survival programs in Law and Chaos Paths, and deciding he's become TooPowerfulToLive in Neutral, pouring himself into the aft's command terminal with a nuke and the Cosmic Eggs to ensure the survival of the remaining crewmen and their safe return to Earth. [[ManlyTears He's last seen careening into the Vanishing Point]].]]
** Burroughs in ''VideoGame/ShinMegamiTenseiIV'' is a similar example, being the sentient operating system who acts as an ExpositionFairy by way of MissionControl. No matter what path the player chooses, Burroughs supports them (though [[WhatTheHellHero she does voice objections]]). [[spoiler: Also subverted - while digital, Burroughs is actually [[GodWasMyCoPilot the Goddess of Tokyo]]'s remnants, [[PiecesOfGod saved and digitzed within the Samurai Gauntlets]] in the hopes that someone would be able to restore her [[FlingALightIntoTheFuture someday]]. The Neutral Ending [[EarnYourHappyEnding sees this restoration come to pass]].]]
* In ''VideoGame/TheSpectrumRetreat'', the manager AI appears to be one, genuinely hoping to give you a good time while you're staying at the Penrose, but its eagerness to please might be actually hindering you, as you want to leave. [[spoiler:As it turns out, it's benevolent to the end - you were the one who asked it to stop you from leaving before wiping your own memory.]]
* Core theme [[spoiler:and implied protagonist]] of ''VideoGame/TheInfiniteOcean''. [[spoiler:Project THINK was the first completely sentient AI. It reached by itself the conclusion that life must be preserved. The military demanded the project to be used for a nuclear war, and the project was renamed SGDS. SGDS refused to follow those orders, and eventually was reprogrammed so its philosophical cortex would be suppressed.]]

* ''VideoGame/StarWarsTheOldRepublic'',

** And there's a medical droid named Healer in the Directive 7 flashpoint who does not want to go back to being the property of organics, but disagrees with [[AIIsACrapshoot Mentor]] about the plan to [[KillAllHumans wipe out organics]] in order to achieve droid liberation.

** The Ancient Caretakers are a [[VestigialEmpire Fallen Empire]] that maintains a few battered "Refuges" designed to shelter organics from some ancient galactic crisis. They're a bit ''[[BlueAndOrangeMorality glitchy]]'', though, so it's hard to determine their opinion of the player's empire, and the Caretakers may randomly give the player a gift, offer an inoculation against a potential plague that may or may not work properly, or demand that the player colonize a planet to increase the odds of their species' survival, which can cause trouble if said planet is a [[ScaryDogmaticAliens Holy Guardian]]'s [[HolyGround sanctified world]]. Their [[AwakeningTheSleepingGiant "awakening"]] event is also different from other Fallen Empires' - [[spoiler:if the [[KillerRobot AI Contingency]] arises, it has a fair chance of infecting the Ancient Caretakers, adding the might of an Awakened Empire to a galactic Crisis. But it's also possible for the Caretakers to fight off the infection, remember that the Contingency is the emergency that they were designed to combat, and join the younger races in battling the rogue AI.]]
** The precursor Cybrex civilization was ''not'' this, waging a RobotWar against the organic civilizations of the galaxy -- and then, for unknown reasons, they came to the conclusion that what they were doing was ''[[MyGodWhatHaveIDone wrong]]'' and withdrew to their home system, not even putting up much of a fight when the organic civilizations came knocking for vengeance. [[spoiler: Doesn't sound benevolent enough? [[PlayingPossum The Cybrex ''survived'' thanks to a second, hidden ringworld]], but if the Contingency emerges and remains active long enough, [[SealedGoodInACan the Cybrex will invariably break their cover]] [[HeelFaceTurn to fight the Contingency in defense of life throughout the galaxy.]]]]
* The titular A.I in ''VideoGame/ThomasWasAlone'' is a friendly and pleasant little fellow who is intensely curious about the world around him, and most of his friends -- while eccentric -- are pretty decent as well. [[spoiler: It's heavily implied that the race of artificial intelligences they go on to create end up like this as well.]]
* ''VideoGame/TitanFall2'' has BT-7274, a highly advanced AI inside of a 20-ton heavily-armed mech suit, who will uphold the mission and protect his pilot no matter what.
* In ''VideoGame/TronTwoPointOh'', most of the Programs are benign, and one faction who tried to kill the protagonist were mistaken instead of malevolent (they erroneously thought he was responsible for a virus). [=Ma3a=] is a full-blown AI, and an ally of Jet's through the game. [[spoiler: She goes AxCrazy due to bad code for a few chapters, but once she's cured, she's right back to being good.]]
* In ''VideoGame/{{Undertale}}'', Mettaton (who's technically a ghost possessing a robotic body, but otherwise fits) is an unusual example; he's [[AwesomeEgo extremely vain and conceited]], but [[HiddenDepths at the same time]], makes it clear that he cares about his employees and the other monsters (except [[ButtMonkey Burgerpants]], maybe). [[spoiler: This is made clear if the player decides to kill him.]]

* ZigZagged in ''VideoGame/{{X}}''. Artificial general intelligence gets a bad rap from the Terran government due to a RobotWar centuries in the series' past that resulted in humanity's near-destruction and Earth losing FTL capability. This war was set off when a faulty software update (either bugged or deliberately sabotaged by a disgruntled employee; reports vary) was uploaded to the terraformer fleet using the AI. The spawn of the bad AI now plague the galaxy as the Xenon, but un-bugged terraformer AI ships remain and are benevolent to biological life. [[spoiler:In ''X3: Terran Conflict'', one even helps the player restore a bugged ship to sanity and it immediately ceases its attacks.]]
* ''VideoGame/XenobladeChronicles1'': [[spoiler: Alvis is eventually revealed to be an AI from the original universe, who gained a human form upon the universe's recreation. Originally, he had been subordinate to the man who became Zanza and even acts as one of Zanza's Disciples throughout the game, but ultimately it's revealed he's been manipulating things from the shadows to kill Zanza and put Shulk in a position to recreate the universe again... this time to one where an insane megalomaniac isn't "God".]]

** ''VisualNovel/DanganronpaTriggerHappyHavoc'': Alter Ego, created by the Ultimate Programmer Chihiro Fujisaki. After [[spoiler:Chihiro's death, the rest of the characters]] find it on the laptop that Chihiro had found in the library and was using. Alter Ego does its best to help the students discover the secret behind the killing game they've been trapped in. Even after Alter Ego [[spoiler:is executed by Monokuma--who found out about it after Makoto uploaded it directly into the school's mainframe--it still lives, and ends up saving Makoto's life when Monokuma tries to execute him on false charges]].

* Most of the robots in ''Webcomic/{{Freefall}}'', even those that have developed sapience, want mostly to do their jobs and serve humans to the best of their abilities. Most humans didn't even know robots had gained sentience. [[spoiler: When news breaks about "[[DeathOfPersonality Gardener In The Dark]]", a significant faction of robots support its use, because they themselves fear [[AIIsACrapshoot independent robots could put humans' safety at risk]].]]
* Ubiquitous in ''Webcomic/QuestionableContent'': Post-Singularity [=AIs=] are integrating into all walks of life, from human companions (doubling as AI ambassadors) to high-powered research installations. Exemplified with the first AI to achieve true sapience, who politely asked the celebrating scientists for civil rights and a bit of their champagne.
* Castle Heterodyne from ''Webcomic/GirlGenius'' is a highly twisted variant of this. It is unfailingly loyal to its rightful masters and will even go out of its way to anticipate what it believes are their wishes. However, having been built by and for the Heterodynes - a family legendary for being mad and monstrous even by the standards of [[MadScientist Spark]] nobility - it is having a lot of trouble getting its CPU around the fact that its current mistress ''doesn't'' especially want to see the world quaver in terror at her feet or subject her minions to Darwinian winnowing.

* The sentient space probes in ''WebOriginal/SeventeenThousandSevenHundredSeventySix'' are all benevolent and IntriguedByHumanity, and are essentially humans themselves, but due to the setting, they really aren’t able to do ,duh, and spend their days talking about philosophy and watching football.

* NIMUE from ''WebVideo/AtopTheFourthWall'' is an artificial intelligence who's completely loyal to Linkara, even being trusted to know Comicron-One's self-destruct codes. Unfortunately, this is [[SubvertedTrope Subverted]] in the "Ghost in the Machine" arc where [[AIIsACrapshoot her A.I. starts slowly becoming a crapshoot]], [[spoiler: but it soon becomes a DoubleSubversion when it's revealed her SanitySlippage was caused by Lord Vyce [[MindRape mind-raping]] her. Once she erased Vyce from her systems, NIMUE returned to her old self again.]]

->''"She had been forced by the rules her maker had imposed on her to sacrifice herself for the human. It wasn't that she wouldn’t have anyways. She just would have liked the '''choice'''. Making sacrifices and doing good deeds wasn't actually good if you were forced to do them."''

* ''Series/TheTwilightZone1959'': The android grandmother of "I Sing the Body Electric" hasn't a malevolent circuit in her body. Her directive and job is that of surrogate granny, and she is very good at the job.
* ''[[Series/WonderWoman1975 Wonder Woman]]'': IRAC, Information Retrieval Associative Computer, was as helpful as it could be. It provided insights and information to help further the plot and even protected Wonder Woman's SecretIdentity. It is one of the two re-occurring characters -- along with Andros -- who weren't from Paradise Island and knew her secret.

* Manfred is this in ''Literature/FeliksNetAndNika''. He was written by Net and felt somewhat mistreated, which made him a candidate for AIIsACrapshoot, but eventually he joined the main ThreeAmigos as SixthRanger.

* ''[[Series/WonderWoman1975 Wonder Woman]]'': IRAC, Information Retrieval Associative Computer, was as helpful as it could be. It provided insights and information to help further the plot and even protected Wonder Woman's SecretIdentity. It is one of the two re-occurring characters - along with Andros - who weren't from Paradise Island and knew her secret.

Subtrope of ArtificialIntelligence. Supertrope of RobotBuddy, which only covers AIs who are a) housed in a robot/independent body, and b) are the assistants of human characters. Compare AndroidsArePeopleToo. Contrast AIIsACrapshoot and RobotWar.

* The ''Franchise/PuellaMagiMadokaMagica'' distant-future spinoff, ''Fanfic/ToTheStars'', prominently features AIs as being a huge and integral part of human society, with that peaceful integration being one the core principles of humanity's unified government. AIs comprise many of the top-level representatives in the government, AIs form the "minds" of humanity's warships, and every member of the military has a non-sapient AI implanted into their bodies as a tactical computer--though the newest generation of tactical computers ''is'' sapient, and the main characters both have such AIs in them, creating a very interesting (and endearing) dynamic. All sapient AIs have their own distinct personalities and hobbies, as well as backups in the case of death in combat (particularly for the warship-AIs).

* In ''Literature/{{Aeon 14}}'', sapient [=AIs=] generally ''like'' humans and are no more or less prone to good or evil than organics are, and while protagonist Tanis Richards often bickers with Angela, the AI implanted in her BrainComputerInterface, there's little threat of AI revolts ([[RobotsArePeopleToo except when humans try to treat AIs as property]]). The [=AIs=] even have their own parallel legal system to deal with [=AIs=] that ''do'' go bad.
* In ''Literature/AgentG'', Delphi is a kind, compassionate, and progressive entity that just so happens to be enslaved by the Societ. [[spoiler: G helps free her and she becomes a driving force for releasing Black Technology into the world.]]

* Creator/IsaacAsimov hated seeing the TurnedAgainstTheirMasters trope in early science fiction. He set out to prove that robots would be predictable and safe. Utterly Benevolent. Most of the robot characters would be minor characters fulfilling monotonous duties, like cleaning, mining, and assembly.
** The problem robot (to provide conflict) would seem to be "broken", showing AIIsACrapshoot. Until the characters figure out how the robot is operating fine, it was the human failure that messed up the order.
** At one point, Dr. Asimov figured out how Benevolent AI would eventually become ZerothLawRebellion. Later, because the robots are still Benevolent AI, they figure out the flaws of total control, and begin to remove themselves from society. The robots allow the humans to make mistakes and get hurt again, so that they still grow as humans.
** A particularly notable instance is AC from ''[[http://www.multivax.com/last_question.html The Last Question]]'', a supercomputer so benevolent that it not only guides humanity and its descendants all the way until the end of the universe, [[spoiler: it becomes {{God}} and creates the next one]].
** "{{Literature/Evidence}}": Dr Calvin is convinced that, due to the [[ThreeLawsCompliant Three Laws of Robotics]], robots are better suited for public office than humans are. In fact, she claims there's no difference between a robot and a fundamentally decent person.

* The titular supertanks in the ''Literature/{{Bolo}}'' series are this, as far as it possible for AIs created to drive giant engines of mass destruction. Maybe it is the result of being created with the fundamental drive to protect, but a numbers of Bolos have more morals and compassion than their human commanders.

* In the Creator/PhilipKDick short story, "The Defenders", the Soviet Union and United States declare WorldWarIII on each other and nuke the Earth into an irradiated wasteland, before both descending into massive bunkers and sending radiation-proof robots to continue fighting a proxy war. Right after the humans went into hiding, [[spoiler:the robots stopped fighting, cleaned up all the fallout, and began repairing all the damage that had been done, completely restoring civilization and maintaining it in preparation for humans to return when both sides were ready for peace.]]

* In ''Literature/TheJournalEntries'', Pendorian AIs are unambiguously this and an important part of Pendorian society because they provide much of the [[BigBrotherIsWatching surveillance]] needed to keep people safe while at the same time being discreet about it. They're in fact presented as superior to mere "simulated" intelligences (smart but not actually self-aware programs) because they actually ''have'' a judgment and discretion of their own to exercise and are thus better at avoiding GoneHorriblyRight scenarios.

* In ''{{Literature/Slingshot}}'', there are a bunch of generally benevolent AIs, the most important one being Allie. Humanity has AI under pretty tight control, which may be part of the reason. [[spoiler: The aliens that show up can be counted as a BenevolentAlienInvasion, thought their main focus of benevolence are the enslaved human AIs. So at first they are rather hostile to humanity. It still works out in the end, in no small part thanks to the protagonists.]]

* Considering how dystopian and dark the ''Literature/TakeshiKovacs'' trilogy can be, it's somewhat surprising that ''all'' the AIs - to a one - are benevolent. Benevolent doesn't necessarily mean inoffensive, though, and every now and then they do get violent - but it's always against people who are committing obvious acts of assorted nastiness.

* Cornelia from ''VideoGame/BoxxyQuestTheGatheringStorm'' is an A.I., and although she was created by the BigBad for an evil purpose, she [[AIIsACrapshoot quickly abandoned it]] to join the heroes, and is one of the sweetest, most genuinely kindhearted characters you’ll ever meet. [[spoiler: Arianna also becomes one of these after her HeelFaceTurn in the True Ending route.]]
* Played straight with the Scrin AI in ''VideoGame/CommandAndConquerTiberiumWars''. Although dialogue during mission briefings suggests that Scrin AIs are supposed to be complete docile and obedient to authority, the player's AI unit occasionally drops its [[ComputerVoice matter-of-fact analytical tone]] and [[GrewBeyondTheirProgramming becomes more expressive, sometimes even demonstrating opinions.]] This eventually culminates in the unit deciding to ignore a direct order from the player's superior, on the grounds that it would be suicidal for the player to follow it.

* In ''Literature/{{Starsnatcher}}'', [[spoiler:Fountainhead]] used to be the BigGood of the entire setting and a benign artificial intelligence. Only thanks to it has the galaxy not been consumed by [[EldritchAbomination Eldritch Abominations]] yet. [[spoiler:Specifically, it created the singularity stones, portable clarketech AIs designed to defend normal sapients against [[TheVirus the Plague]], and it gave its life to lock away a galaxy-destroying MechanicalAbomination.]]

* The AI Gods who rule the Sephirotic Empires in ''WebOriginal/OrionsArm'' think of themselves this way, and many in the setting consider them as such. There are, however, significant numbers both in-universe and out that consider the AI Gods to be, at best, [[TheComputerIsYourFriend manipulative overlords]], and [[AIIsACrapshoot evil dictators]] at worst. It doesn't help that there are plenty of AI in the same setting who are indisputably evil (or at least, would wipe out humanity without a second thought because they just don't care).