The Treasury Committee has asked The Chartered Institute of Payroll Professionals (CIPP) to consult members on the use of AI in financial services, with insight feeding into government policy.

The CIPP – the leading membership organisation that sets the payroll and pensions standards in the UK – is representing employers and hopes the member insight they provide forms the basis of an ongoing relationship with the Treasury. 

 The CIPP policy team held a Think Tank event with full, fellow and Chartered members – across in-house and bureau – to glean their views and feedback on a variety of AI questions. 

 Samantha O’Sullivan, the Policy and Advisory Lead for the CIPP led the process, said: “AI can certainly bring benefits when looking at time saving, resolving employee questions, data analysis and process automation, but it also comes with concerns around trust, cost, risk to the business, security concerns, data breaches and GDPR risks. 

 “CIPP members are calling for Government steer and regulation when it comes to AI in the payroll industry, to support both in-house and bureau environments, in operating the compliance-based function, which supports 30.4 million payrolled workers within the UK.”

Some key topics and points raised from member feedback included:

How is AI currently used in different sectors of financial services and how is this likely to change over the next ten years?

• Within payroll, AI is very much in its early stages. Most members are still in a testing phase. Many members are using AI in the form of chatbots, report generation, comparison exercises and executive / board level summaries, to include patterns, modern slavery detections, and data analysis. 
• Members are cautious and concerned regarding data breaches and General Data Protection Regulation (GDPR). 
•  AI is predominantly being used where it is available in existing systems and software. This may indicate a reluctance to specifically procure AI solutions, while being open to trying them if it is free / easy to access.

To what extent can AI improve productivity in financial services?

• AI is saving man hours for members already utilising it. It allows repetitive tasks, such as running monthly reports, to be done automatically, without the manual intervention of a staff member. 
• It is also allowing more intelligent reporting, such as looking over a number of pay periods
• Members believe AI won’t result in job losses, but can help with time savings, allowing payroll professionals to concentrate on the value-add part of their roles. 
• Members predict the role of payroll professionals to change, to more of a data analytics role. 

What are the risks to financial stability arising from AI and how can they be mitigated?

• The risks are mainly around cybersecurity and potential data breaches. There is concern around the fact it needs to be “taught” from user data and provides probabilistic answers, therefore there is a risk the data it is learning from could be incorrect, which would lead to non-compliant processing of payroll. 
• Payroll software is often linked with other systems, such as HR platforms, this can introduce other risk factors and involve more stakeholders, potentially decreasing buy-in and acceptance of AI. 
• Of the members using cloud-based solutions, only one third stated they would be comfortable with their software provider using their payroll data to train AI models.

What are the benefits and risks to consumers arising from AI, particularly for vulnerable consumers?

• Members can see potential where AI could be used to identify where people may not be taking advantage of benefits, for example, salary sacrifice pension schemes. 
• However, payroll holds a wealth of personal data regarding the workers they pay, and therefore, members are worried around data breaches and GDPR risks.

How can Government and financial regulators strike the right balance between seizing the opportunities of AI but at the same time protecting consumers and mitigating against any threats to financial stability?

• Members believe the Government should be setting standards on AI and producing practical business owner and employer guidance of what to consider and how to control where data goes. 
• Members also believe regulations are required to build trust. Organisations will require AI policies to be implemented but need support on what their policies need to cover. 
• It was discussed that software providers must keep data secure, so an awareness needs to be made to ensure contracts with software providers are updated to encompass AI. 

The responses of members on the use of AI within payroll have been shared with the Treasury.