Slashdot Mirror
Thousands in London Face Incorrect Benefit Cuts From Automated Fraud Detector (sky.com)
Thousands of people could soon be receiving letters threatening to cut off vital housing benefits as they face being incorrectly targeted by a new automated fraud detector. From a report: The government-backed London Counter Fraud Hub, developed by BAE, has been hailed a success after being trialled in four boroughs - Camden, Ealing, Croydon and Islington. Using vast quantities of data from millions of households, it is designed to target potential fraud cases involving the single person council tax discount, subletting in local authority housing and business rate relief and rating.
Ealing, the lead council for the project, found the automated elements of the system targeting single person discount fraud was 80% effective -- which is seen as an acceptable benchmark. With just over one million claimants of council tax single person discount in London, the London Counter Fraud Hub estimates it will detect around 40,000 fraudulent cases in the first year. Critics say the 20% error rate is unacceptable as around 8,000 people will receive letters wrongly accusing them of fraud.
Ealing, the lead council for the project, found the automated elements of the system targeting single person discount fraud was 80% effective -- which is seen as an acceptable benchmark. With just over one million claimants of council tax single person discount in London, the London Counter Fraud Hub estimates it will detect around 40,000 fraudulent cases in the first year. Critics say the 20% error rate is unacceptable as around 8,000 people will receive letters wrongly accusing them of fraud.
Don't bring it to America or it'll be pealing like crazy around this Dunnald Drumpf traitor's White House.
Nothing wrong with using an automated system that assumes conclusions. As long as you have
1) A viable system in place for remedying the failures. If you're unsure about the soft word in there, peg it at "better than youtube's"
2) The claimed conclusions are asserted tentatively. (The data in our system suggests) YOU LIED ABOUT X AND Z WILL HAPPEN (if not addressed).
Unfortunately we fuck up both. Privatize the savings, socialize the fuck up, charge forward trusting the data without hesitation, why stop at flags and alerts when you can have it execute outright.
If people want "benefits" let them get job and fucking PAY FOR THEM like all the rest of us always do. Vast majority of people who get them are not really even need them. Like all welfare queen they just prefer to leech off goverment and keep voting for other leftists to keep the gravy flowing at taxpayer expenses. The fraud algoithm detected them BECAUSE they are all the frauds!
80% efficiency is ok if the other 20% are false negatives. But false positives??? What???
They do need to work on their error rate, but they should be working on the false negatives. The false positives can be handled with a human investigator.
John McAfee 'It was like that time I hired that Bangkok prostitute; to do my taxes, while I fucked my accountant'
Sometimes I wonder if builders of automated systems ever bother considering false positives and the consequences. It's almost as if they collect statistics and show, "Hey, our system has a 98% success rate detecting fraudulent this-or-that!" and ignore "Ooops, it turns out it just randomly flagged 98% of this-or-tat as fraudlent, and thus caught 98% of real fraud at the expense of flagging 98% of legit this-or-that incorrectly as fraud."
"To err is human; to really foul things up you need a computer."
Table-ized A.I.
...did the same basic thing. It's about saving money in two ways: fraud reduction (ostensibly) and people being wrongly kicked off and basically giving up rather than spend hours on phone calls that achieve nothing (or get terminated for no reason), waiting at offices to deal with overworked and capricious staff only to be told "you need to do that online" and dealing with an alternatively confusing and a-functional web interface. Also popular with parts of the electorate who enjoy watching the poor getting blamed for predicament.
Automated fraud detection is fine.
But acting on a number which is basically a statistical guess, without some follow-up manual review is plain irresponsible.
The credit industry does this and regulation differs by country. For public services, the justice standards should be different.
I don't think we are hearing the whole story though. It is hard to believe that a 20% error rate is accepted by European standards of fairness.
I've always been bothered that all police in the US carry guns - where in general the police in the UK don't, at lest in the England of my youth. But make no mistake, there is a pernicious truth: despite unarmed Bobbies, England is a police state. Just look at the speed control on the roads. Variable Speed limits with long distance tracking, not a person in sight. I know there is a safety argument, but it is absolute control over people's movement. Literally. But that's not my point.
The real story is how a 20% Error Rate is even close to acceptable in any automated system, especially in an environment where addressing issues through bureaucracy is practically impossible?!?! On top of that every run will yield a new chance to be falsely accused so the threat always looms. Seeing this in my own USA is a much deeper fear for me than any terrorist activity, and 40% of the public will welcome it, including people who will be ruined by it. It's insane.
This is part of the thick edge of the wedge. Beware all.
We don't need $30 aspirin and $250/ 15min doctor bills.
Organize and let your MPs see the homeless caravan migrating away from the city towards a plastic tent village in the forest. Remember to look miserable on the road as the police bothers you and make sure to take all interview requests. You can do it during summer. Change the system through shame and death!
benefits cliffs just make people work less or not work at all.
Why work MC's part time as if you end up working 22 hours one week vs say just 20 you lose benefits that take a long time to get back.
It is all a disaster. If you are in hospital dying we tell the local council that you are in hospital therefore they should not send bailiffs round to evict them. Because they are dying they are entitled to "housing benefit" more than half of them receive threatening letters of eviction. The same with the elderly depending on how old they are if they are in their 80s they will be OK if they are in their 70s, they will receive threatening letters. We do not let housing associations or local council officers onto the ward of terminally ill people. We pin their letters up sealed for social services to collect. When they die we have a "click to tell all" of the person's death with an agreement of a member of the family this stops all bills and direct debits, which sends the bank and properties into probate, allowing the bank to issue an emergency payment to a member of the family to pay for the funeral, and make a claim for the rest of the money in the bank with a death certificate as proof.
If "tell all" is not agreed by a member of the family then the DWP will make a claim for money back and demand a copy of the death certificate. If you have been taking money before the tell all you can be prosecuted for theft.
Well, actually, you have to look at the entire million anyway. Because you want to know the false negatives too. You just run through the 40k flagged first, and you review them before you send out any accusations.
Looks like they made numerous errors along the way, starting with failing to discern true/false negative/positive, precision and recall, and all that. That makes this an exercise in political posturing. From an upper class already renowned for shitting on the lower classes. Good show, guise.
I do hope the UK government has permission to use this data in this way. Otherwise they could be fined up to 4% of the GRoos National Product.
And here everyone was so afraid automation was going to take all the jobs! This is going to force that review, and its going to be people doing those reviews. They are going to need a lot of caseworkers with a 20% error rate flagging accounts falsely!
Yer jerbs are safe, Ingaland!
I thought they said frog detection.
Have gnu, will travel.
Critics say the 20% error rate is unacceptable as around 8,000 people will receive letters wrongly accusing them of fraud.
And if you say Zero then you have to prove to me that the prior system had zero errors.
Demanding perfection in replacing a flawed system is ridiculous.
I guarantee there is a appeals process after detection - people will not be thrown out of their homes instantly.
Ken
To think, a government organization is actually using machine learning this way. Humanity has officially become an idiocracy.
If the fraud is then MANUALLY checked by somebody competent in the same department, fine. if on the other hand this lead to an automatic guilt process, or requesting the person to prove their "innocence" that begins to be far far less fine. And that seem to be partially the case as those "fraud detection" recipient are sent automatic letters. How would you feel if your government accused you of fraud in an official letter ?
What is the error rate of their existing fraud detection?
It's all those Pollacks, Ruminants and Albino Huns. They'll all be sent packing soon thanks to that lovely Mr Farage.
He's so dapper, isn't he? Never see him with without a proper shirt and tie.
You couldn't get bananas. They all went to the brave boys in the trenches you know. Unless you dropped your knickers for any Yank that walked by, that is. So I heard.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
Does this look like this article is about a 'benefit being lost'(That is to say money you are given) or is this about a 'tax deduction being revoked' (that is to say money you don't have to pay)?
Also, the seriousness of the issue, would seem like it has a lot to do with how difficult it is to prove the system wrong. I mean if the notes being sent basically say , call and prove you are how you say you are, or live where you say you live, etc. ( and that can be done reasonably by sending some receipts somewhere and it takes maybe 2 hours ). How much harm is actually being done. I mean, they could just ask EVERYBODY to prove their eligibility EVERY Year, as it now uncommon for taxes or programs in the United states, but this seems easier.
Oh well, at least they don't have to worry about an Indian jet bombing their mud hut.
It's acceptable because they can force a 'guilty until proven innocent' argument. Ideally, the disadvantaged will sue the government for defamation and lack of due process. Alas, that won't be happening.
they pulled the same shit in australia with welfare, throwing yearly data into an algorithm that divides by 26 to get fortnightly ammounts, problem is, it doesn't account for if you finished using welfare partway through a working year, so people have been hit up for repayments they don't owe the government in the first place!
fraud from govts is on the rise.