Video games for Yu-Gi-Oh! have a particularly poor track record in this area.
While some of the games' idiotic moves can be justified by the fact that the AI couldn't possibly know the identity of your facedown cards, and that the kind of analysis that would allow a player to even make the right guesses can be really difficult even for human players, some of the cases are a little more obviously Artificial Stupidity.
In some games the AI will use an effect that requires paying life points when they have that exact amount of life points left. For example: AI has 800 life points, AI plays Premature Burial, AI pays 800 life points to use Premature Burial, AI immediately loses.
Then you have Mokuba, for whom this trope is invoked intentionally. What a digital dummy! To give you the idea of how dumb he is, his second strongest monster is Kanan The Swordmistress, a normal monster with 1400 ATK and 1400 DEF. He summons none of his monsters in defense mode, letting you just keep knocking them down. His entire strategy is to draw one monster, Cyber Stein, which has the ability to summon a fusion monster. This is the only way you can lose to him, because if he manage to do this, he'll summon Blue-Eyes Ultimate Dragon. Which he will gleefully prompt to crash and burn into your obvious Mirror Force/Sakuretsu Armor/Maneater Bug/Penguin Soldier/etc. Or even worse, into your more obvious Magic Cylinder, which will cause him to kill himself without failnote Activating Cyber Stein costs 5000 life points. Magic Cylinder causes an attack by your opponent's monster to be cancelled and the ATK of the monster to be substracted from his life points. BEUD has 4500 ATK. A normal match has a player begin with 8000 life points.
In many of the earlier games, such as Eternal Duelist Soul, at harder levels, the AI essentially knew the ATK and DEF of any of your facedown monsters, and would make its decisions whether or not to attack based on that. Some of the "good" duelists like Yami Yugi go at you with cards that technically can destroy yours in battle...and then leaves them right open to a strong counterattack when the player is able to capitalize on the fact that they left a monster with 1000-1100 ATK in attack mode at the end of their turn. Attack! Attack! Attack! meets Artificial Stupidity here.
The AI in Tag Force 2 is considered one of the worst examples of this in a Yu-Gi-Oh game, to the point where it seems like the game is actively trying to sabotage your efforts when you play a tag duel. A clear example came from a Tag Force 4 (video here), when the AI used Prideful Roar against Clear Vice Dragon. For the unaware: Prideful Roar pays LP to increase a Monster's ATK to be slightly higher than its target's, but Clear Vice Dragon's ATK is always double that of its attack target ...so it increases again when Prideful Roar activates. The AI paid 2800 Life, took more than double that in damage, and promptly lost.
The AI in general seems to have trouble with Monsters that can increase or lower ATK. For instance, say you've got Psychic Commander, a 1400 ATK monster that can lower the ATK of Monsters it battles, and the AI has Mystic Tomato, which has equal ATK and can Summon a weak Dark monster when destroyed. The AI is programmed to ram "searchers" into Monsters with equal ATK, so the AI will attempt to ram Mystic Tomato into Psychic Commander, then when that doesn't work, it'll Summon another Tomato from its Deck and do the exact same thing. It will repeat the process until it runs out of Tomatoes. One of the most infamous cases of Tag Force partner stupidity was this video. In a nutshell: The opponent, Para, had Suijin on the field. Suijin can, once per Duel, reduce the ATK of an attacking Monster to 0, so if you try attacking it, you'll lose your Monster and take a ton of damage in the process. Three guesses what the partner, Bastion, did, and the first two don't count. And Bastion is supposed to be The Smart Guy...
Tag Force Special is the first game to feature Pendulum Summoning, and the AI simply does not know how to handle it. Yuya is by far the worst in this regard, with a near-insane aversion to putting his Pendulum Magicians in their proper spots and a refusal to use Entermate Wizard's effect. On top of this, the AI is set to always perform a strategy when possible, even when it's not a good idea. Pegasus is the most obvious offender, as this Duel shows. And even when they do have a workable strategy, they often don't use it; Kaiba will often Summon Kaibaman in ATK and then do nothing with it, even when he has a Blue-Eyes right there in his hand.
Tag Force Special also reveals a rather large bug in the game's AI, since one of the Starter Villains from Yu-Gi-Oh! ARC-V, Eita, uses Ordeal of a Traveler. It makes your opponent guess the type of card in your hand whenever they try to attack, and if they fail to guess, their attacking monster gets bounced. The main trick with the card is remembering the cards in your opponent's hand... something that the AI simply does not do. Even when Eita has one card in his hand, he's revealed it to be a monster four times, and he hasn't played any other cards, your tag partner will continue to rush in, guess "Spell Card!", and get all your cards bounced.
In Yu-Gi-Oh! Dark Duel Stories, the AIs have a bad habit of offering high-ATK monsters as tributes to summon something of the same strength or even weaker, example: Offering "Jirai Gumo"(2200ATK/100DEF) as a tribute to Tribute Summon "Catapult Turtle" (1000ATK/2000DEF). The AI also likes to use monsters who have lower ATK than DEF to attack, as long as the ATK is at least half the DEF. Sometimes, Yami Yugi will use "Megamorph" (which acts like a universal Equip card, increasing a monster's ATK and DEF by 500) on Mystical Elf just so that he can attack... with 1300 ATK.
There's also its inability to judge the worth of cards in its hands, meaning that it discards randomly whenever an effect makes them do so, which can often make them cripple their entire strategy by eliminating their most important card. To wit: The AI has three cards, which consist of a weak monster, a strong monster whose level is too high to be summoned, and a spell which makes the user discard a card but would let him summon the stronger monster. The AI will, 50% of the time, activate the spell, discard the stronger monster, and then summon the weaker monster which wouldn't need the spell in the first place.
Yu-Gi-Oh! 5D's's Duel Transer/Master of the Cards is also not immune. The AI Computer opponent you have unlocked initially has a few decks that are easy to overcome, but for some reason it likes to set off a combo of Waboku and Hallowed Life Barrier. Waboku stops you taking damage that turn and stops your monsters from being killed, Hallowed Life Barrier is basically the same, except you need to discard a card to activate it, and all it does is nullify battle and effect damage, not protect monsters.
The AI is incapable of deciding whether or not using particular traps is a good idea or not. If your opponent has Torrential Tribute set (a trap which wipes all monsters on the field when activated), they'll use it even if the monster they already have on the field is stronger than the one you just summoned (of course if you're doing this, they might foresee your equipping it with something). Then again, they'll often wipe the whole field even if they have a much stronger monster out. Opponents using Torrential Tribute to destroy the whole field when they have a 2500+ ATK ritual monster out when all you did was summon a relatively weak monster is common enough to count as a strategy to get rid of their monsters.
Not to mention that they will tribute summon their powerful monster, and then play Torrential Tribute, wiping out everything. Also done with Dark Hole.
Despite being the main character, Yugi will often make the baffling decision to keep summoning Sinister Serpent, an effect monster with 300 ATK and 250 DEF. It's effect is to keep showing up in his hand if it's destroyed. Good if you plan on sacrificing it, but he never does this. He keeps it out until you vaporize it with a much stronger monster, and then keep summoning it just because.
Total Defence Shogun is particularly weak in the hands of the AI. It has 1550 ATK, 2500 DEF, and it can attack while in defence mode. Whenever they play/use/control one however, they will always switch it to attack mode. So, basically, the AI weakens the monster by 950 points, AND opens themselves up to Life Point damage voluntarilly.
The AI will sometimes use Premature Burial or Call of the Haunted to summon Gearfried the Iron Knight. Either of those cards can be used to summon a monster from the Graveyard, but the card is then equipped to the monster; if the card is destroyed, so is the monster it summoned. Gearfried destroys any card that is equipped to it automatically. Yeah... Even more humorous because Premature Burial costs 800 life points to use.
The AI has also been known to do things like take control of your monster using a card like Change Of Heart, which takes yours for one turn, but then boost its stats with a permanent equip spell. So at the end of your turn, you get your monster back, only the AI has actually helped you. Similarly, it's a common player strategy to Suicide Attack a weaker Monster into a stronger one, because the weaker Monster has an effect that activates in the Graveyard (see: Sangan, Mystic Tomato). The AI will do this with Monsters they've taken control of, even though cards revert to their owner's control in the Graveyard - so if an AI attacks with a stolen Sangan, it goes back to your Graveyard and YOU get the effect while they take damage. Thanks, buddy!
In Yugioh: Dungeon Dice Monsters, any character not found in the anime will just summon around their Heart Points and will eventually use up all their summons. They will then be unable to do anything, allowing you to have as many rolls as you need to summon anything. The Exodia pieces can be summoned this way, and by summoning them all, you get an instant win, and the AI is powerless to stop you. You can beat anyone in the game with an equally inane strategy. There are summonable "items" in the game which take the form of chests. Only the summoner knows what's in the chest, and it activates when a monster passes over it. The AI will never run over your chests, in the expectation that it might be a trap (and, to be fair, it might). However, it is possible, by spamming cheap summons, to block your opponent so that the only path to your heart points is through the chest. At which point, the AI will helpfully sit around, waiting for you to kill them. Also, the AI will never attack your monsters unless it can one-shot-kill the monster, and usually not even then. The AI will never use its own items even if they're beneficial, so you can use their items to power up your monsters or evenrevive your destroyed monsters. The AI will never use its monsters' special effects (the sole exception being Orgoth the Relentless) even if they have the crests to do so. You can be right next to the enemy's Die Master getting ready to secure the winning strike, and the AI still will not attack your monster or attempt to protect itself. Again- with rare exceptions. The AI will also head straight for your die master and get stuck in corners and dead-ends as a result. The AI will even select dice that make it physically impossible for the AI to summon a monster.
Macro Cosmos Decks turn most games against the AI into a comedy of errors. Macro Cosmos removes any card that would enter the Graveyard from play, meaning that many strategies centering around the Graveyard become fairly crippled. A human player would attempt to destroy Macro Cosmos as soon as possible, then initiate their normal strategies. Not the AI, which will completely ignore Macro Cosmos and continue to play as if it wasn't there. You haven't seen Artificial Stupidity until you've seen an AI use Foolish Burial to remove its own Treeborn Frog.
Oftentimes, there'll be at least one character who plays an Exodia Deck. Exodia is an infamous set of cards: five weak Monsters that, if you hold all five in your hand, let you instantly win the Duel. A human player will try to protect the pieces at any cost, not Summoning or discarding them unless they have a way of getting them back - if you lose a piece, it's game over, since Exodia Decks usually feature little beyond drawing cards and maybe defensive cards. The AI has no such qualms, perceiving them as only weak monsters, and will regularly ditch the pieces at the first opportunity. The worst part is when you've got a clear field and the AI decides to summon a 200-ATK Left Arm of the Forbidden One to attack you.
Yu-Gi-Oh! Reshef of Destruction mainly averts this, but the computer will always attack your cards if they are face down in defense mode unless their monsters have 0 Attack. This will happen even if you use a card to cover up previously seen monsters.
An app called Yu-Gi-Oh! Duel Generation generally has incredible AI., but Pot of Duality seems to cause some hiccups. Instead of taking something useful like Mirror Force, they'll take Skelengel. They also take a Field Spell card when they already have the exact same card already on the field instead of a monster to defend their life points.
Yu-Gi-Oh! Destiny Board Traveller has this when the Outer Space board is used. The AI will heavily prioritize summoning monsters, and generally ignore the special spaces, which is a very bad idea on this board, since if a player doesn't go to a special space for 5 turns, they will lose 1000 Life Points. It is very possible to win on this board with this being the only way your AI opponents take any damage at all.
The final boss of Magic: The Gathering: Battlegrounds has the ability to cast any spell in the game, any time he likes. Theoretically this means he should be able to spam you with giant monsters while countering any spell that you try to cast. Instead, he just sort of hangs around not doing much, and can be trapped in a loop by summoning the same low-level Mook over and over again. Possibly intentional on the part of the developers, since if the boss used his powers in a sensible fashion then he would be completely unbeatable.
Sometimes, the computer can come up with masterful combos and expert tactical plans. Other times: they sacrifice their last point of life to Pestilence in order to kill some Llanowar Elves, and summoning a Lord of the Pit and then doing nothing with it, meaning it eats all the computer's monsters and starts on the computer's life total. In particular, it will only attack if the creature is guaranteed to survive the creatures you have out or it has enough monsters to zerg you to death. This means that it doesn't, for example, fling expendable creatures at you to whittle down your forces, even if those creatures have a significant upkeep like sacrificing a creature.
Pestilence itself is a global enchantment that deals 1 damage to EVERYTHING ON THE FIELD (including BOTH players) for 1 black mana, repeated as long as you have black mana/creatures (Pestilence is destroyed if no creatures are on the field). Lord of the Pit is a creature with 7 power (deals 7 damage), 7 toughness (can take 7 damage), is flying (can only be blocked by other fliers or creatures with "Web"), and has FIRST STRIKE (His 7 damage can be spread however he likes first, and then any defenders still alive deal damage to him). This came at the drawback of needing to sacrifice a creature every round, if you didn't, he did 7 damage to his controller. Many a game was won by simply wiping the creatures away, or stalling until your opponent ran out.
Force of Nature had 8 Power, 8 Toughness, and had Trample (and extra damage above and beyond a defending creature's toughness hit the player-usually extra damage is wasted). This was at the expense of 4 green mana every round, or Force of Nature did 8 damage to the player. Destroying land, however, is tough, but one card made it easier: Living Lands. Living Lands turns all Forests (which PRODUCE the vital green mana) into 1 power, 1 toughness creatures. The AI was more then happy to fling these pipsqueaks at you. You can see where this is going...Extra Stupid in that AI decks seemed to run both Force of Nature and Living Lands together. Only one deck had Force of Nature without Living Lands, and that one's a Green/White mix deck.
This has plagued computerised Go engines (especially when compared with computerised Chess engines), with them being trounced by professional Go players even when given 25 stone advantages... The latest Go AI can win with a 9 stone advantage, and has been stated that it's up to good amateur levels. In Go the problem space is much larger. While both go and chess have a finite number of moves per turn, determining the possible moves in chess is a matter of thinking of each piece and seeing where they can land and if it's open, whereas in go it's not a matter of "which of these 32 pieces can move where?" so much as "which of these 300-odd spots should pieces go on?", which doesn't just make calculation slower and more memory intensive, but also makes the heuristics harder to work on, too.
A classic computer game that has gone by many names over the years relies on this trope. In the original version, you had to run from robots, although modern versions have used zombies, vampires, Eldritch Abominations... basically, whatever. Anyway, you and the robots both move one square per turn (like a chess king), and robots will chase you down. You have no weapon, but the robots will attack and annihilate each other before they ever turn on you! Thus, you have to rely on robots' tendency to kill each other before they kill you. It's even been done with Daleks.
The Windows program Mission Maker has extremely primitive AI. Make a character 'Seek and Destroy' the player, then get another character between them. The hostile character, instead of moving around, will kill the other character to get to the player.
Exploiting the Artificial Stupidity of the guards in Lode Runner is very useful, with some levels relying on it. For instance, you can position yourself on a ladder so they climb upwards when you're directly below them.
In Splinter Cell: Conviction, at one point you are confronted with an enemy helicopter gunship. It always shoots in front of Sam and never thinks to try and flank him.
The usual method to beat the last boss in Guitar Hero III invokes this. Basically, there's a certain point where a Whammy attack will kill him in one hit. Why is this? In that particular section, instead of using the whammy bar to recover, he just hammers the STRUM BAR until he kills himself. One critical flaw in an otherwise complete bastard.
Computer controlled helpers in Kirby Super Star have their uses, but don't expect them to live very long. Fortunately they're easy to replace. Also, your AI allies in Kirby & the Amazing Mirror can be counted on for jack squat. They'll mill around in random areas, getting random abilities (including abilities not in their current area), and if you call them to your side...well, it's usually for one of three reasons: a boss fight, the fact that they bring health-restoring food with them, or one of them somehow snagged the Smashability and you're just waiting for them to screw up so you can use it yourself.
If you've ever played a video game adaptation of a game show, you've probably encountered computer contestants that couldn't answer simple questions correctly. Press Your Luck for the Wii is one of the egregious examples, with computer opponents answering questions such as "What animal do we get milk from?", "What is 36 divided by 6?", or "How many months are in a year?" wrong.
Old PC or video game versions of Jeopardy! in the early 1990s had the AI contestants buzz in and answer in complete gibberish. The answer pool was so small that pulling a wrong answer from that could clue another player in later. Other versions had no answer pool, resulting in the correct answer or the same gibberish every time. Examples include ZWXYZ on the Game Boy and XXX on the Genesis versions. This is true for the NES versions as well (save for Super Jeopardy!), but the gibberish is the exact same length as the correct response, and often shows some letters in the response as well. For example, if a correct response is TVTropes, the AI would show something like *V@r#pes.
Demigod, an early MOBA, tends to inflict this on players when they go against the bots on higher difficulty levels. The opposing team will specialise in hit-and-run tactics, prioritise game-changers like Reinforcement Flags, and just generally give you a run for your money. Your allies, on the other hand, will position themselves directly between two enemy gun posts and pick on irrelevant minions, while being whaled on by the enemy, thus feeding your opponents both gold and experience. Since your opponents are now relatively stronger, and can afford to upgrade their defensive structures, this process becomes streamlined, resulting in ally deaths roughly every few minutes.
The Star Fox series has wingmen's "calling for a help" as a fixed pattern in every side scrolling stages. They can't help themselves and will go down if you don't help them. All-Range Mode, however, turns their stupidity up to eleven. One particularly notable example of how bad the wingmen's AI is in All-Range Mode is in the Star Wolf dogfights in Star Fox 64. Each Star Wolf pilot is programmed to target a specific member of your squadron. Each wingman will constantly plead for you to help him by shooting down the Star Wolf member who's on his tail. Once you do, he will blissfully fly around in a circle minding his own business and make no effort to help you as the remaining Star Wolf members continue to rip you and your other wingmen to shreds. In a way, this is a blessing, as you don't get points if your wingmen score the killing blow on something. In Sector Z, they become rather competent when it comes to taking down missiles, which is not a good thing for medal huntersnote and no, you can't just shoot down your wingmen; if any of them are down, you cannot earn a medal.
In The Simpsons: Road Rage, buses constantly crash into anything in sight without any provoking them, typically you.
Wheatley, also known as the Intelligence Dampening Sphere in Portal 2 is a deliberate In-Universe example, described by GLaDOS as "the product of the greatest minds of a generation working together with the express purpose of building the dumbest moron who ever lived", and "the moron they built to make me an idiot". It's actually not a totally straight example, as this A.I. was made to be like a stupid human (and programmed very well for that purpose), rather than simply a badly programmed computer.
One animated board game for MS-DOS called Hexxagon was indeed a lot of fun. Pit red gems against chrome drops on a hexagonal board in deep space. Landing next to your opponent's pieces would transform them into your own. The hard Craniac was usually pretty good, but when it was running out of spaces to go, it tended to make stupid moves such as jumping pieces into spaces it could have cloned into and in the process of doing so, often opening up holes allowing its opponent to capture some of its pieces. These stupid moves usually cropped up when the Craniac was losing, so it rarely changed who won the game, although if you had been narrowly losing to it, such a move could turn the tide in your favour. On lower difficulties, the Craniac also tended to make stupid moves much more frequently, but in those cases, it was expected behaviour.
In many electronic versions of Monopoly, AI opponents are programmed to accept trades based on the value of the properties offered, rather than the situation. (Ex. getting an AI player to give you the third Orange property in exchange for a single Green property). While a human player can exploit this shamelessly, it also works when the AIs trade with each other, leaving you watching helplessly as one AI trades Boardwalk for a light blue, a purple and $25.
Akinator can be surprisingly daft at times, since he doesn't really understand the questions he's asking (they're all user-contributed); as a result, he may keep asking you almost identical questions, or directly opposite ones ("Is your character real?" followed by "Is your character fictional (does not really exist)?") Of course, he'll ask if they have black hair and then ask if they have blond hair, plus the always entertaining "Asks if they're from one universe, is told yes, guesses a different universe". This is likely the only program to think Mass Effect and Halo are the same universe. And sometimes it will ask you if you've met/said hello to the character after you've confirmed it's a fictional character or dead historical figure. Apparently it believes you might be Thursday Next or a time traveler.
The Bot Wars expansion for the Battlestations board game implements this in a tabletop game. The player's opponents are rogue robots which begin the game with below-average skills, and can learn as the adventure progresses. However, the bots also have a huge mothership; ship size is a factor in the difficulty of piloting, the mothership is the highest size possible in the game, and the bots' skills are below average.. No matter how the GM runs the bots, unless he/she fudges the dice rolls, it is perfectly possible for the bots to wreck their own mothership in the opening scenario due to their incompetence in piloting it.
The autoplay option of most (if not all) solitaire games falls into this; the computer will always move a card onto the foundation piles (and thus out of play) as soon as it can, even doing so is the worst choice possible.
Your wingmen in Aero Fighters Assault are generally more competent than those in Star Fox 64 (probably because the levels are so big and the draw distance so lousy that you'll almost never see them and they can kill enemies without actually having to kill enemies) but they're just as bad in regards to frequently asking for help and being generally helpless. They also will not even attempt to attack the stage bosses, and destroying the bosses is the one thing you need to do to finish 6 of the 8 levels. A harmless- though amusing- example in the dialogue: Volk will often follow up his "thanks for helping me" line with his enemy taunt, leading to him saying "Much thanks, Comrade; now you will die!" A more meta example in that the two levels where you have no wingmen have a significantly better frame rate- and thus much better controls- than the other levels of the game. The extra strain placed on the game by creating and keeping track of your wingmen slows the game down painfully in the bulk of the levels.
Computer Scrabble games on the hardest level have the entire dictionary available to them, meaning they will have more bingos (using all seven letters for (word score +50 points)) and highest-scoring words in general, however it absolutely stinks at strategy, opening up triple word scores for you and leaving vowels next to triple letter scores (consonants are worth more points) and not playing in areas that would block the human.
In Gruntz, this applies to both your gruntz and the enemy's. Your gruntz will walk in a straight line towards their destination, carelessly walking into any obstacle that is in their path. And both your gruntz and the enemy gruntz can be easily distracted by toys given to them, no matter what happens around them. If the toy is a mobile one, then it gets even better, since they will randomly drive into any instant-kill obstacle near them. On multiplayer levels, the enemy shows some remarkable amounts of stupidity, such as:
Walking straight into pyramids that are about to raise (instant kill), under rolling boulders (instant kill), onto bridges that are about to sink (instant kill)...
If they have giant straws and stand on a goo puddle, then they will glitch and just stand on the puddle instead of walking next to it to drink it.
They will break any bricks standing on their way to your gruntz, even if it means destroying their own fortress' defense.
The AI for Talisman: Digital Edition ends up having fairly frequent moments where it shows that it is a different kind of special:
It will happily overload its inventory by crafting a raft at a woods space, then take a talisman from a player that attacks it before their next turn (despite already having one) and drop the raft it just spent the previous turn crafting, and then on its very next turn, craft another raft.
It will always drop the Amulet from its inventory even if it just picked it up off the ground, even if it is a character that usually doesn't have enough Craft to even hold on to a single spell.
It will always pick up every item it encounters that it legally can, including the previously mentioned amulet. When this puts it over the inventory limit, it may discard a much more useful item (such as a weapon)
The interactive game Fašade can be beaten entirely by answering "yes" to every question, as proven by Brutalmoose in his video.
In Wonderland AdventuresNPCs will make no effort to avoid danger when you tell them where to go, and will always take the same path between tiles even if it kills them.
The first video game based on the Wipeout television series has downright idiotic AI. There are only a small handful of obstacles in the entire game that the AI is capable of navigating. Bump the difficulty past the easiest level and the AI will complete one obstacle each in the first and final rounds and wipe out on everything else through zero interference on your part. A heaping dose of The Computer Is a Cheating Bastard is used to offset these tendencies.
In Absolute Despair Girls, the Monokuma enemies have a problem at getting around walls chasing after dancing Siren Monokumas. They just run around trying to get to it, but failing.
Enemies in the original Trine are not clever when it comes to water, which kills them instantly, and falls, which do likewise if of sufficient distance. There's even one boss-level enemy - the second crystal troll you face - which can be lured into charging over a cliff edge, leading to instant death. The goblins in the sequel are a bit smarter, although they're still not good with environmental hazards or not shooting each other.
You can choose the intelligence of the AIs at the start of the game along with your difficulty; the three options being Silly, Normal, or Smart. When Silly is selected the AI is very easy to beat and poses little challenge to even new players. The ability to choose their intelligence means that you can make your CPU teammate very competent at the Smart setting while making the enemy team stupid, assuring that you'll win most of the minigames. However, if the player is winning by too much of a margin then (depending on the difficulty setting) the enemy AIs will act smarter and start getting luckier.
Ms. Pac-Man: Maze Madness has this in the multiplayer mode.
In Dot Mania mode (collect 80 Pac-Dots first to win), AIs have a tendency to run into ghosts and hazards even if there's an alternate path that they can take.
In Da Bomb mode (a 30-second bomb is held by one player, who must pass it to another; be the last one alive to win), AIs often make dumb decisions, like running to the back in a straight line into a wall after tagging a living player, thus making them easier to get re-tagged again.
In general, when they're not blatantly cheating when it comes to the rules, the multiplayer AIs tend to run back and forth or use portals for no reason. Both cases can leave them defenseless to ghosts and hazards.
The AI in Hitman (2016) does not always handle hazards well. As seen in thisAchievement Hunter video, they will assume that the hazard has been neutralized (which is normally true), even if the corpses of dozens of other people making the same mistake are clearly visible in the hazard. (To say nothing of how nobody thinks twice about stepping in the hazard in the first place...)