Labour: Embrace AI to reform welfare services and help more people into work

A Labour DWP will be a ‘trailblazer’ in the use of AI, data, digital and tech to help people find jobs, tackle fraud and improve welfare services.

Speaking at the Social Market Foundation today [Tuesday 11 July] Jonathan Ashworth, the Shadow Work and Pensions Secretary will outline his ambition to fully embrace technology an AI to reform DWP services helping jobseekers move into quality jobs faster.  

Ashworth will say that 13 years of Conservative government has left the critical functions of the Department for Work and Pensions in a state of disarray. He will point to figures that show:

• Nearly 40 per cent of claims are processed late,

• £8.3 billion has been lost to fraud and error in the past year

• There were over 126,000 disputes about errors regarding real times earnings raised by claimants in 2022.

• Each year, billions in social security entitlements including nearly £2 billion of pension credit for older people is left unclaimed.

Criticising the Government record, the shadow work and pensions secretary will say that a Labour government will modernise and reform welfare services, prioritising three areas for better use of technology.

• Improved use of AI technology in Jobcentre Plus to personalise return to work plans, making full use of job-matching AI platforms to help jobseekers find suitable work.

• Better use data science and artificial intelligence to process claims faster and drive take up.

• Going further with machine learning algorithms and AI to tackle fraud and error currently costing £8 billion a year.

Jonathan Ashworth MP, Labour’s Shadow Work and Pensions Secretary, will say:

“Key to our mission to hit the highest growth in the G7, tame inflation and raise living standards is helping more people move into good, quality well paid jobs. 

“As AI transforms the labour market opening new opportunities in the future of work so too can it help deliver a new phase of reform across DWP services to better support people into work and help firms fill vacancies quicker.

“The creation of Jobcentre Plus was a key public service reform of the last Labour government, the next Labour government will pioneer new reforms to better equip jobseekers meet the challenges of the labour market of the future.

“Utilising powerful new AI tools will help Jobcentres offer more personalised, tailored support for jobseekers helping them move into employment faster with a job that matches their skill set.

“AI in the hands of Jobcentre work coaches will better help jobseekers with CV analysis tools and interview practice preparation. When we know AI is now routinely used for rapidly assessing CV applications for current jobs. Let’s now make it focused on getting people the skills they need for future jobs as well.

“Better use of data could ensure real-time analysis of local labour markets helping Jobcentres better understand local conditions and respond appropriately when arranging suitable coaching and training for jobseekers.

“Just as we can help people into good jobs we can, through the better use of data, also help vulnerable citizens access payments they are entitled to and currently not claiming or receiving late.

“And the shocking amounts of fraud and error overseen and disgracefully considered acceptable by this Conservative government – over £8 billion on most recent figures - can be confronted making more use of machine learning and pattern detection techniques.

“For too long the DWP has been offering an analogue service in an AI age.

“Labour is the party of welfare reform. A trailblazing Labour DWP will embrace technology and AI helping us meet our priorities on getting more people back to work and grow the economy.”

Ends

Notes

Jonathan Ashworth will be speaking to the Social Market Foundation think tank at 1130am on Tuesday 11 January.

Examples of the use of technology that could be utilised include:

• Personalised support to help people re-enter the job market faster, and  with a job that matches their skill set. This will include:
• AI job matching schemes to better link job seekers with suitable jobs
• AI CV analysis to suggest improvements and identify next career steps (including training or work) and job interview practice
• AI recommended training courses to better support Jobcentre work coaches to find and arrange appropriate training.
• Real-time analysis of labour-market demand based on open vacancies, how quickly they are filled, job requirement skill descriptions and geography, to ensure Jobcentres better arrange suitable coaching
• Machine learning algorithms to detect potential instances of fraud by searching for anomalies to help reduce fraud and error

Nearly 40 per cent of claims to the DWP are processed late

https://www.gov.uk/government/publications/dwp-annual-report-and-accounts-2022-to-2023

There were 126,000 disputes about errors in 2022

        https://automatinguc.co.uk/sites/automatinguc/files/assets/Answers to Benefit levels inquiry_Final_2.pdf

There is £2bn of unclaimed pension credit

        https://www.gov.uk/government/news/eligible-pensioners-urged-to-claim-pension-credit-to-help-with-cost-of-living

The UK is currently falling behind in the use of AI to transform services:

France: The French public employment service, Pôle Emploi, uses a tool called “Automatic CV Analysis” (Analyse Automatique de CV – AACV) which helps jobseekers from the moment they create a profile with Pôle Emploi. Based on the CV uploaded by the jobseeker, AACV suggests skills that the jobseeker is likely to possess but did not mention on their CV. The case worker can validate or reject the suggestions made by the tool, following an interview with the jobseeker.

Belgium: In Flanders, the public employment service VDAB uses the Jobbereik application to help users visualise the set of occupations and jobs that they could consider given their existing skills and competencies. Users enter their desired occupation to receive a list of possible alternative occupations that require more or less similar skills profiles. Given that many people focus only on the occupation last exercised when searching for a new job, Jobbereik may open up new employment opportunities by highlighting occupations that they would otherwise not have considered.

Finland: Finland’s Centre for Pensions utilised AI to identify 80% of retirees who will retire on a disability pension two years before they retire. The ML algorithm was tested on the Centre’s database of 500,000 people, and correctly identified 78% of retirees who would claim disability pensions. 

Estonia: Piloting the use of AI in job centres: Estonia introduced a system that analysed data about an individual’s circumstances (skills, last job, time since becoming unemployed) to provide job coaches with suggestions on what support would be most effective, based on what has worked best for similar individuals in the past.