Woman becomes first person in the UK to win legal battle using AI law firm
A woman has become the first person in the UK to win a legal battle using an AI "law firm."
The healthcare worker, who had a stellar performance record, felt helpless when bosses placed her on a Performance Improvement Plan. Unable to afford a solicitor, she turned to Grapple Law – the UK’s only legal practice for individuals fully operated by bots. In just a few weeks, her case was resolved and she won £30,000 – without a tribunal or human lawyer in sight.
In the 1960s my father worked in what is now called IT when computers were huge machines tended by engineers permanently on site. Decades ago he predicted that computers would one day do the work of lawyers.
And here we are.
16 comments:
And not only lawyers but anyone working with a laptop or pc can be replaced in three years. Doesn't of course mean they will be replaced. Fund management houses are replacing back-office people, AI can reconcile trades and calculate NAVs much more accurately and quickly. AI doesn't get sick, need maternity leave, ask for a raise, go on a holiday or go on strike.
Decades ago I predicted that computers would replace investment managers. A colleague disagreed vociferously.
I inferred that his father may well have been an investment manager.
Anyway, the rise of ETFs suggests I was right.
My daughter is an accountant. Daddy is quietly suggesting she diversify.......
The question that has to be asked is did anyone actually check that the legal arguments being made by the AI lawyer are real or not? The fact the bot won this case does not mean it was actually putting forward genuine legal arguments, or quoting actual laws and legal precedents. It might have been putting forward a pack of AI 'hallucinations' that no-one bothered to check whether they were genuine or not.
Question is ... is this frightening or encouraging?
Does the AI need to 'know' any law. Perhaps all that's necessary is to send a convincing sounding letter from a 'lawyer'. The recipient thinks, 'oh dear, they are going to put up a fight' and backs down.
Anon - yes, it's time to assume that glitches will be sorted and something major is coming. Even if AI turns out to be weaker than the hype, it already looks impressive.
dearieme - yes ETFs do suggest you were right - leave it to the computer.
Chris - it's not easy to advise the younger generation, but working at a desk does seem to be vulnerable.
Sobers - it's a guess, but this feels like a supervised trial run with the aim of getting the system off the ground and pulling in some publicity at the same time. If so, then presumably its performance will have been analysed afterwards.
James - both I reckon.
djc - this may be all that it did because I believe AI report and letter writing systems are already used by lawyers.
I suppose the next question is how long before two (different) AI's sue each other?
Dave - they may be doing it already, issuing thousands of pages of arguments to make their case.
Take Walmart for instance. Their staff turnover annually is 75% ie they walk out and replacements need to be trained etc. 2.1mn employees. If 10% of them was replaced by robots (Optimus), they would need 60,000 robots to do the same work. Annual savings would be $8bn.
They wouldn't even have to fire anyone - 75% of employees just walk out every year!
Anon - interesting, supermarkets must be looking at robots for some jobs and if that works they will take it further. I believe Ocado have a robotic selection system for their deliveries.
They do. And they sell their system to customers all over the world!
Tammly - there's a video somewhere too, must look it up.
Andrew Orlowski isn't that keen on AI:
https://www.telegraph.co.uk/business/2025/12/15/world-grip-ai-mania-consequences-devastating/
https://www.youtube.com/watch?v=ssZ_8cqfBlE
How many robots does it take to run a grocery store?
Dave - yes he has been a sceptic for a long time and maybe AI will meet some barrier it can't cross. One barrier may be cost, it turns out to be too expensive and the return on AI investment is inadequate. Or there may be something about human intelligence it can't emulate. Or it may work.
djc - a fascinating video, thanks. Is it many robots or one robot with many working parts?
Post a Comment