Artificial intelligence is increasingly being used by benefits claimants to maximise handouts, The Mail on Sunday can reveal.
Programmes such as OpenAI‘s ChatGPT are being used to generate answers and write forms for claimants – contributing to a welfare budget which is predicted to hit £100 billion a year by 2030.
It comes as the Department for Work and Pensions (DWP) has also deployed AI tools to help streamline decisions and decide whether individuals are eligible for benefits.
This newspaper’s investigation found that across social media, claimants are boasting about AI securing benefits payments.
One individual said a bot had been her ‘saviour’ for ‘doing the rest for you’ after instructing it to produce template forms for her.
‘Thank me later’ for this advice, she said. Others share the best ‘prompts’ for the chatbots to produce better results on a range of different benefits.
One user named ‘Mr PIP Hero’ has created his own ‘AI chatbot that fights for your PIP [Personal Independence Payment]’ – which is said to be ‘built to act like a KC in disability law’.
Online forums discussing AI usage for PIP payments are also rife, with users bragging ‘extensive’ uses to create ‘accurate templates’ tailored to different conditions.
Programmes such as OpenAI ‘s ChatGPT are being used to generate answers and write forms for claimants – contributing to a welfare budget which is predicted to hit £100 billion a year by 2030
In one case, a disability organisation has designed a tool for ChatGPT called the ‘Personal Independence Payment form filler’ which will write forms for you.
Disability Horizons is an online shop which helps ‘enhance your life’. It designed the chatbot, which completes applications and provides ‘proactive suggestions’ for forms, to ‘increase your chances of success’.
It claims the chatbot is a ‘stress-free way’ to access benefits.
Last week official statistics showed more than 4 million people are receiving jobless benefits – the highest ever recorded.
Meanwhile the ‘risk-based verification’ AI tool the DWP has developed assesses whether welfare applications are genuine.
Officials believe it will help save £1.6 billion by 2030. But critics say the widespread use of AI could vastly undermine this.
Welfare assessments are done in person, by video link or over the phone, but the most recent figures show only 5 per cent are currently conducted face-to-face.
Returning to face-to-face sessions would make it more difficult for people to fake or exaggerate conditions than when using chatbots, but there is no indication this will happen and the Government has resisted calls to enforce it.
