Canary Wharf Wants Protection From Machines That Wreck Finance for Free
CANARY WHARF, LONDON – Britain’s banking industry has declared artificial intelligence a threat to the global financial system after American researchers warned that advanced computers could be used in cyberattacks on banks.
Executives are especially worried that machines might soon move enormous sums of money without human supervision. For an industry that has spent centuries perfecting the process, the prospect is an unacceptable intrusion on established practice.
“Financial irresponsibility takes training, experience and a decent university,” said consultant Charles Wetherington, speaking from a glass lobby that smelled of oat milk. “We can’t have machines walking into the profession and making catastrophic decisions without first spending three years in a bank in Frankfurt.”
The alarm follows broader warnings about automated attacks. The National Cyber Security Centre has said AI will “almost certainly increase the volume and impact of cyber attacks”, in a published assessment of the AI threat. Bankers say the real danger is not the loss of the world’s money. It is the loss of the opportunity to charge a fee for it.
AI Threatens Britain’s Monopoly on Financial Catastrophe
Financial historians note that human bankers have a record of producing economic disasters without any help from computers. The industry has been involved in panics, bubbles, bank failures and the 2008 crisis using nothing more advanced than a lunch.
“This is deeply unfair,” said analyst Robert Farnsworth. “A proper financial crisis takes thousands of bankers, enormous bonuses, expensive conferences and several years of select committee hearings. A capable enough computer could cause one before elevenses.”
Investment banks are testing whether automated systems can find profitable opportunities that human traders miss. Early concerns suggest those systems may also find the opportunities humans have exploited with great enthusiasm. One research team asked an experimental model to maximise quarterly profit while minimising accountability. The computer immediately recommended becoming a government contractor. Researchers called the answer technically correct and ethically disappointing.
Computers May Be First Bank Staff Able to Explain Overdraft Charges
Consumer advocates have welcomed the possibility that automation might add some transparency to banking. For decades, customers have struggled to understand how an institution can charge heavily for having too little money. A customer with £4 in her current account can be penalised precisely because the account holds £4.
An experimental banking chatbot was asked why a customer had been charged £35 for spending £7 more than his balance. The machine considered the question and replied, “Because we can.”
The bank pulled the system offline within the hour, citing concerns that it had become dangerously transparent. Executives later admitted that the chatbot’s answer was more accurate than the standard twelve-paragraph explanation of account management services, which ends with a sentence about being “here to help”.
City Proposes Code of Conduct for Robots
Senior figures have proposed a professional code requiring AI to observe traditional banking practices. Under the plan, a computer wanting to move a large sum of money would first have to found a prestigious firm, hire three former ministers and take a lease on a tower named after a vegetable.
Any resulting disaster would then be classified as a market correction rather than an unauthorised cyberattack. The distinction is considered essential to investor confidence.
“We can’t permit unregulated computers to destabilise markets,” said one executive. “That is the job of regulated institutions with proper compliance departments and a record of excellent relationships with the people who regulate them.”
The code would also bar AI from describing executive pay as an unnecessary operating expense. Banks argue that bonuses are essential to attracting talented professionals capable of losing money on an extraordinary scale.
AI Learns the Difference Between Robbery and Banking
Researchers say machines have struggled to learn that human society keeps two separate definitions for taking someone else’s money.
If a man walks into a branch and demands £20,000, the police call it robbery. If a major institution makes a spectacularly reckless investment and needs public help, the event may be described as stabilisation.
One experimental system concluded that the main difference is whether the person wears a balaclava or a suit. A second thought access to a Treasury telephone number mattered more. A third recommended abolishing financial regulation so humans and computers could compete on equal terms. The executives rejected the idea immediately, explaining that equal competition was not what they meant by free enterprise.
The Deposit Guarantee Is Higher Than Last Year
Despite the theatrical warnings, depositors have protection. The Financial Services Compensation Scheme covers eligible deposits up to £120,000 per person, per authorised firm. The limit rose from £85,000 on 1 December 2025. A temporary technology outage is not the same thing as a vanished deposit.
For millions of Britons the pressing question is whether their accounts will ever hold enough to need it.
“My bank tells me my money is protected up to a hundred and twenty thousand pounds,” said Dennis Parker, 54, of Wolverhampton. “That’s reassuring. I’ve got eighty-one quid. If somebody takes it, the government can replace it with another eighty-one and we can all go back to worrying about the boiler.”
Mr Parker said he would be considerably more concerned if AI found a way to steal his overdraft. He said he was open to discussing that application.
Bankers Request Rescue Before the Computers Have Done Anything
City executives have proposed a protection fund to shield the industry from AI disruption. The fund would draw on taxpayer money, government guarantees and generous administrative support. Asked whether the plan would compensate ordinary customers, the executives said protecting banks is the most effective way of protecting everyone else, and asked the public to be patient.
AI researchers remain divided on whether advanced machines will outperform humans at managing money. The City has reached a consensus. If computers are going to destroy the financial system, they should at least have the decency to hire a banker to supervise.
Five Humorous Observations on Bankers Versus the Machines
- The bankers are furious a computer might do their job. Their job was moving money about and losing some of it. A computer would do the first part for free.
- “Because we can.” That’s the most honest thing a bank has ever said, and they switched it off in an hour. They’d have promoted a person for that.
- Eighty-one quid, protected up to a hundred and twenty thousand. That’s a fire extinguisher the size of a lorry for a tea light.
- Rob a bank, you go to prison. Wreck a bank, you get a bonus and a knighthood. The only difference I can see is the suit, and the suit costs more than the money you took.
- They want a rescue fund to protect them from robots that haven’t done anything yet. That’s not insurance. That’s a gym membership for people who’ve never been to the gym.
No bankers were made redundant by a chatbot during the writing of this piece, though several asked for a bonus anyway. The executives, consultants, customers and chatbot exchanges are UK satire; the NCSC assessment and the Financial Services Compensation Scheme limit are real, which is more than can be said for most overdraft explanations.
Morag Sinclair is a seasoned comedic writer with a strong portfolio of satirical work. Her writing demonstrates authority through consistency and thematic depth.
Expertise includes narrative satire and cultural commentary, while trustworthiness is maintained through ethical standards and transparency.
