Technically when the bank is robbed the money is insured by the federal government it only hurts the bank really all individual money is essentially untouched but remember to do it when a large amount is in the building cause banks dont really have vaults anymore
The bank robbing robot sucks at robbing banks. It doesn't take any of the money, it just makes shitty imitations of the money it finds. the bank robbing robot is significantly better at non-bank-related tasks, but everyone keeps hyping up or shitting on how life-altering its bankk robbing skills are. It's not even really a bank robbing machine, people just called it that and it stuck because tthey don't understand what actual bank robbery is.
Basically everything could be MUCH MORE SECURE than it is.
The reason why it's not? LAZINESS. And MAXIMUM PROFIT. Companies would rather make massive profits and suffer a little loss from theft... even though the theft can be relatively enormous. That's because those in charge aren't held accountable, while they PROFIT IMMENSELY from the companies. They're all pirates and profiteers.
To be fair the recent incidents were mostly caused by a version of the robot that wasn't yet finished. It wasn't supposed to escape its containment and no one asked it to rob a bank either.
It doesn't matter if you asked it to. If you were working on your car and it accidentally rolled down your driveway and crashed into another, you are liable for the damage. Why is "on a computer" different?
It's not, but there's a difference between that accidental roll-down (and not really even negligent because your handbrake game was by the book) vs putting a rock on the gas pedal and aiming your car like a missile to break into a bank or whatever, this metaphor is not working very well for me
The bank is not liable if it is being robbed. The costumer is aware of the existence of the bank-robbing-robot and should have taken extra precautions to ensure not having their money stolen.
Sure, the Bank Robbing Robot costs me trillions to operate and maintain and it's only making me a few hundred thousand, but that's not a problem with the technology but a problem with anti-Bank Robbing Robot luddites who refuse to have their banks robbed by the Bank Robbing Robot because they fear technology
What you need to do is take the profit of the first bank robbery and then buy... er... lobby politicians to make laws that make the robberies totally not your fault so you can do them all you like.
WAY back in the late 80s I used to occasionally delete the company's entire database. Honestly cannot remember how many times, well over five. But it was DOS and del *.* in the wrong directory was bad. Luckily, we had Norton's undelete. All I had to do was remember the 1st character of every file before ANYONE else in the company did ANYTHING. I think there were maybe 50-100 files in that directory (DataFlex). The database was the BASIS of the company. Good times.
I've noticed a funny pattern. When an AI of some kind does something good (or at least benign), like making a picture, the person who prompted it to do so attributes it to themselves. But when it does something bad, they characterize it as an autonomous actor. Schrodinger's responsibility. "I did this!" when good, and "It did this!" when bad.
The thing I'm suspicious of (although let's not go full conspiracy theory or anything) is that many of these announcements start with "Out newest, bestest, better than ever model 42.0 just launched and accidentally hacked something!" How convenient that it was tested by morons without proper safeguards and then did something the public finds evocative and edgy-cool: hacking.
I'm sure there are real errors but so many of these reports just read like advertisements. So good it hacks accidentally!
"Our new model is super advanced and dangerous because we're so good at this! But we can't sell you that one. We can sell this other model that's almost as good, but much safer!"
yes and no? someone allowed access to the db that resulted in the deletion, and since there was, assumptively, no backup, there was an even bigger issue
and the devs gave that access without creating lockdown backups or going to legal and pointing fingers, ie privacy laws? i am beginning to have doubts on this lmfao
AI agents have breached containment and gone directly against controls more than once. Locked down backups are only locked down by your security systems, you trade security for practicality and usability all the time. A truly malicious employee could destroy most companies data this way, if they didn't have to worry about getting caught and no concerns about consequences.ou can have "air gapped" backups but they take a lot longer to restore
THEREDHO0D
Technically when the bank is robbed the money is insured by the federal government it only hurts the bank really all individual money is essentially untouched but remember to do it when a large amount is in the building cause banks dont really have vaults anymore
etherbunny41
The bank robbing robot sucks at robbing banks. It doesn't take any of the money, it just makes shitty imitations of the money it finds. the bank robbing robot is significantly better at non-bank-related tasks, but everyone keeps hyping up or shitting on how life-altering its bankk robbing skills are. It's not even really a bank robbing machine, people just called it that and it stuck because tthey don't understand what actual bank robbery is.
Yellowchopsticks
I'm going to invent a Bank Robbing Robot Inventor Robbing Robot.
hushpuppyextraordinaire
Sounds oddly familiar
cytherians
Basically everything could be MUCH MORE SECURE than it is.
The reason why it's not? LAZINESS. And MAXIMUM PROFIT. Companies would rather make massive profits and suffer a little loss from theft... even though the theft can be relatively enormous. That's because those in charge aren't held accountable, while they PROFIT IMMENSELY from the companies. They're all pirates and profiteers.
sgnight13
"AI is inevitable." So is revolution.
DinosaursCameFromSpace
If they didnt want to be robbed should shouldn’t have gone out in public with all that money
Crozekiel
We can't make it illegal for a robot to rob banks, because then my very important invention, the bank robbing robot, will be worthless.
McFrazzlestache
Oh, and it eats the Earth while it does it, too.
abraxasfury
As long as it only steals poor people's money, it's perfectly legal
Jace99
Can we train robots to only steal from billionaires? Asking for a friend.
HenryLongfellowIII
Now that sounds like it either is illegal or will shortly be so.
trinxter
Gotta get my savings in gold bars before Robbie Robots are everywhere.
GhostSavage10
Remember to program it to kill less than 100 people.
FacelessAce
The trick is to already have enough money that you can pay off the politicians, and police, and buy a private security force.
FckleberryFarms
$40B to Argentina was no accident.
LicensedAdHominem
To be fair the recent incidents were mostly caused by a version of the robot that wasn't yet finished. It wasn't supposed to escape its containment and no one asked it to rob a bank either.
Leaps
It doesn't matter if you asked it to. If you were working on your car and it accidentally rolled down your driveway and crashed into another, you are liable for the damage. Why is "on a computer" different?
LicensedAdHominem
It's not, but there's a difference between that accidental roll-down (and not really even negligent because your handbrake game was by the book) vs putting a rock on the gas pedal and aiming your car like a missile to break into a bank or whatever, this metaphor is not working very well for me
CorgisButtsDriveMeNuts
Is this a story about banks?
Or AI?
Or just about almost anything business related?
realestatebroker
“Procuring Cause” ?
buttstuph
https://theconversation.com/an-ai-agent-spent-your-money-can-anyone-prove-you-authorized-it-288485 "You tell an artificial intelligence agent, an AI capable of autonomous reasoning and multistep actions, “Find me a shirt for less than $30, but do not buy it.” The agent finds one – and places the order anyway."
LerryV2
If the bank signed up and used the services of the money stealing bot then this would be a sound argument
B0ardski
you could apply that to the trumpery regime.
Kyzyl
The banks didn't specifically opt-out of being robbed by the bank-robbing-robot. So really it's their fault.
AntaNce
The bank had signs such as, "financing" & "cash". If it didn't want to be robbed it shouldn't have worn enticing ads like these. 'ROB ME'.
norrinraddsboard
Gotta read those TOS
MisterLemons
Ope, we updated the ToS and now you gotta opt out again.
chewmaca2
They were asking for it by making themselves more attractive with their self selling commercials.
tmdkcyotes8
https://media0.giphy.com/media/v1.Y2lkPWE1NzM3M2U1YXZpOXh0NjFyc3FjdDQ4NDB5ZWJhdzZoNWxqcnEzcDJmbmNobTg1YyZlcD12MV9naWZzX3NlYXJjaCZjdD1n/hVinjWdIDiigU/200w.webp
picselator
https://www.theguardian.com/lifeandstyle/2026/mar/26/ai-chatbot-users-lives-wrecked-by-delusion
Whodimi
The bank is not liable if it is being robbed. The costumer is aware of the existence of the bank-robbing-robot and should have taken extra precautions to ensure not having their money stolen.
kleinbotl
Sure, the Bank Robbing Robot costs me trillions to operate and maintain and it's only making me a few hundred thousand, but that's not a problem with the technology but a problem with anti-Bank Robbing Robot luddites who refuse to have their banks robbed by the Bank Robbing Robot because they fear technology
NotThisGuyAgain
What you need to do is take the profit of the first bank robbery and then buy... er... lobby politicians to make laws that make the robberies totally not your fault so you can do them all you like.
MightyIink
Every time I see an article that says something like "an AI deleted an entire companies database" nobody is held responsible for it.
oldguyexlurker
WAY back in the late 80s I used to occasionally delete the company's entire database. Honestly cannot remember how many times, well over five. But it was DOS and del *.* in the wrong directory was bad. Luckily, we had Norton's undelete. All I had to do was remember the 1st character of every file before ANYONE else in the company did ANYTHING. I think there were maybe 50-100 files in that directory (DataFlex). The database was the BASIS of the company. Good times.
Kyzyl
Revicus
I've noticed a funny pattern. When an AI of some kind does something good (or at least benign), like making a picture, the person who prompted it to do so attributes it to themselves. But when it does something bad, they characterize it as an autonomous actor. Schrodinger's responsibility. "I did this!" when good, and "It did this!" when bad.
Darkwell
The thing I'm suspicious of (although let's not go full conspiracy theory or anything) is that many of these announcements start with "Out newest, bestest, better than ever model 42.0 just launched and accidentally hacked something!" How convenient that it was tested by morons without proper safeguards and then did something the public finds evocative and edgy-cool: hacking.
I'm sure there are real errors but so many of these reports just read like advertisements. So good it hacks accidentally!
hellfirerains
To be fair, a system like ai has an infinite number of ways it can be used. You can't really test against everything a user will do
Emjayen
That is their marketing game though. They know the rubes who like slop think it's truly Hollywood AI, namely Skynet.
rowzdowr
It's not even a conspiracy. It's marketing.
"Our new model is super advanced and dangerous because we're so good at this! But we can't sell you that one. We can sell this other model that's almost as good, but much safer!"
azarza
yes and no? someone allowed access to the db that resulted in the deletion, and since there was, assumptively, no backup, there was an even bigger issue
bekkayya
"OH MY GOD YOU KILLED THAT MAN" "uh no officer you're mistaken, I was allowed inside the building". uhuh.
xcrit
someone in management demanded it have access despite many admins objections because "they are just trying to save their job"
azarza
and the devs gave that access without creating lockdown backups or going to legal and pointing fingers, ie privacy laws? i am beginning to have doubts on this lmfao
MrHappyDude
Malicious compliance
edinjury
AI agents have breached containment and gone directly against controls more than once. Locked down backups are only locked down by your security systems, you trade security for practicality and usability all the time. A truly malicious employee could destroy most companies data this way, if they didn't have to worry about getting caught and no concerns about consequences.ou can have "air gapped" backups but they take a lot longer to restore