Will AI replace compliance professionals?

Written by Simone Jones on Tuesday 22 October, 2019

The question of whether technology and automation will replace humans in the workplace isn’t new. Today we talk of artificial intelligence (AI) and robots, but the threat posed to workers by technology goes much further back – when Henry Ford first introduced the automated assembly line, it was met with suspicion from workers who feared losing their jobs [1].

The gradual acceptance of former innovations like Ford’s assembly line doesn’t mean we should be nonchalant about new technology – far from it. Rapid tech developments are changing the way we work, and the studies are concerning: a survey from IBM suggests [2] that in the next three years alone, 120 million workers worldwide will need to be reskilled due to the impact of AI.

Another study by PwC [3] found that ‘financial services jobs could be relatively vulnerable to automation in the shorter term’, due in part to workers spending a large amount of time doing ‘simple computational tasks’. Barclay Simpson announced in its compliance jobs market report [4] that 58% of candidates are unsure of how RegTech will affect job security.

AI is today being adopted across all sectors. The Serious Fraud Office (SFO) has used AI to automate document analysis, and as part of its case against Rolls-Royce used a ‘robo-lawyer’ to reportedly scan documents for legal professional privilege content which was 2,000 times faster than human lawyers [5]. The opportunities presented by AI and machine learning to ‘learn’ or adapt to new inputs, recognise patterns and quickly sift through large amounts of data are deeply attractive to businesses, particularly when looking to improve manual processes.

Seizing opportunities

It is clear then that tech encroachment in the work place isn’t a topic for the future but has already arrived. In September 2019 HSBC announced that it implemented a new transaction monitoring tool utilising network intelligence and automated contextual monitoring within its trade finance division. [6]

Traditionally, automated transaction monitoring has been seen as a difficult task to implement in trade finance due to the document-heavy nature of the processes, but by utilising new technology it has now been proven possible. HSBC also announced [7] the use of a financial crime transaction monitoring tool to identify commercial opportunities, not only highlighting ‘red flags’ of potentially unusual activity but ‘green flags’ for new business.

Using AI to anticipate a customer’s need may be new to banking, but it has already long infiltrated the most mundane landscapes of our everyday lives, from Netflix recommendations to Gmail filtering emails into groups. [8]

AI utilised in trivial processes – such as movie recommendations – aren’t required to stand up to a high-level of scrutiny, but going forward, as AI is adopted into more complex processes, the need to understand the rationale behind decisions made or informed by machines will become absolutely vital.

Explaining why AI reached a decision, particularly to consumers that may have had a loan rejected, is a topic that has been explored by the UK’s regulator. [9] Whilst the very nature of machine learning makes full explanation difficult, ‘sufficient explainability’ is the target.  

Understanding the ethics associated with AI is also a compelling broader issue to be confronted, with wider implications than just for financial services and compliance.

The notion of ‘machines’ may bring forth ideas of impartiality free from human bias, but the truth is far more complex. The logic and rules behind such systems were created by people, and with that comes the creator’s ideas and biases.

AI will create opportunities for new roles – compliance professionals will increasingly be required to be conversant with colleagues who work in technology to understand this growing area.

Evolving workforce

The financial crisis and the increased focus (and enforcement) from regulators on financial crime risk saw a bump in resources in compliance teams around the world, including headcount. As we return to a more business-as-usual way of operating, firms will be looking at optimising resources – and this could include using new technology to automate manual processes.

Compliance has been relatively immune from most headcount reductions seen across the financial services sector, but this doesn’t mean that we should rest on our laurels. We must consider in what ways the environment in which we operate is developing.

That will require us to adapt and obtain new skills. People will continue to be needed, but the types of jobs that they do will change. In its report, PwC outlined that in the longer term, those most at risk of losing jobs from automation could be workers without a higher-education, which emphasises the importance of ‘increased investment in lifelong learning and retraining’.

IBM also found that executives were seeking more soft skills from their workforce; flexibility, agility and adaptability to change were the most sought-after competencies.