Business > Efficiency

AI: the long-term solution to public sector IT efficiency?

Neil Merrett Published 13 June 2017

As new research finds some public sector bodies looking to adopt AI technology for operational efficiency, local government is bracing for a longer-term, cautious approach to innovation


New BT research of 1,501 IT decision makers in the public and private sector has concluded that of a third of individuals surveyed that were found to be planning Artificial Intelligence (AI) investment over the next two years, 62% expect more effective operations as a result.

But how viable are these expectations in the shorter-term and what understanding is there of the predicted impacts of increased automation and AI use?

Based on local government experience, particularly for leading digital figures working within London boroughs such as Haringey and Enfield, even if the potential for AI and automated technologies is ultimately overhyped, there is expected to be significant disruption nonetheless.

However, in the short to medium-term, some public sector stakeholders argue there may be limited gains in terms of direct benefits from investing and using the technology to try and drastically overhaul service delivery.

BT’s latest findings considered the current impact of disruptive technologies in the workplace and potential challenges from automated systems and AI, particularly regarding the need for more robust cyber security planning.

Among key conclusions, 71% of organisations that operate within the public sector announced they were using 'big data' to better inform efficiency work, as opposed to 58% of groups operating in the private sector.

Of the same survey group, 44% of public sector respondents said automated processes were expected to make their organisations more open to attack. 22% of private sector decision makers held the same fear, according to the findings.

“The sensitivity of government and public data which organisations operating in the public sector are privy to may have an impact on why these organisations demonstrated a greater concern over implementing these technologies in comparison to those in the private sector,” said the findings.

But how are these findings being reflected in how local government is looking to transform its operations to tackle reduced budgets with disruptive technologies?

Speaking at the 2017 Government Computing Conference in April, Nadira Hussain, head of ICT and transformation at Enfield Council, said the authority was looking at developing its Amelia artificial intelligence platform from IPsoft to support management of user queries.

As opposed to looking at machine learning and robotics as a means to curb employee costs, Hussain argued that the authority’s focus for investing in the technology was to have “a cognitive responsive agent” to help with transactions to free up staff time from having to perform more menial tasks.

“These are things that are run of the mill and mundane processes that customers come to us for,” she said.

Although the focus had been delayed, Hussain argued that the council aims to go-live with Amelia from this summer.  The launch is expected to be a more gradual launch as opposed to going live as a mature, ‘big bang’ solution for transforming council services.

“There is specifically some more work to do to make sure it is relevant and appropriate.  We are going to start with a small proof of concept, probably with some key planning processes and then we will learn from the responsiveness, appropriateness and customer feedback,” added Hussain.

These factors were expected to support an iterative, ongoing improvement process. 

Hussain said that there was strong potential to do away with routine processes in handling requests. However there was a need and preference for retaining human interaction, particularly in the health sector, that was likely to remain indefinitely.

“I am going to be watching with keenness as to how we adopt.  I will certainly want to share those lessons learnt and make them widely available. But what I have understood is there are other public sector organisations and authorities across London that are also expressing an interest in terms of what Amelia might do and how we can use AI with our customers," she said.

Richard Grice, deputy chief operating officer with Haringey Council, said at the event that he found himself alternatively scared and excited by the prospects of AI.

“AI will massively overpromise and will under deliver, but if it under delivers even by 50% it is still going to be massively disruptive,” he said.

Grice expected the technology to have a profound effect on a lot of local government functions, but there was a lack of certainty on the exact nature of what these impacts might be.

“Lots of us are watching Enfield very closely to see what they are doing,” he said.  “One of the interesting things about it is that if you look across the globe and at other industries. I think the figure is something like that 70% of organisations to have significantly invested in AI have done so in their back office and not front office.”

Grice agreed that it was the more routine, transactional tasks that were likely to be the most effective focuses for AI investment.  But other speakers at the conference argued that hopes for an immediate fix to cost or efficiency pressures were likely to be wide of the mark.

Gary Barnett, head of enterprise advisory for GlobalData, expected that in the short to medium-term that there would be some disappointment to the impacts AI will bring to systems as opposed to what he claimed were “absurd promises” made about machine learning,

“But as is often the case, in the long-run, [AI] will surpass our wildest craziest dreams,” he said.

Barnett argued that the initial benefits of the technology were overpromised in the short-term, with there being little clue as to where the technology may lead in the longer-term or what overall impacts it may have on authorities and work.

He said that the best available technology using AI now available could find correlations within large sets of data, but within the last five years alone, there had been huge, in some cases, unthinkable advances in what AI can do.

“In the short-term, AI is about the mundane back office jobs, robotic process automation and frankly that is dross we should have automated years ago,” added Barnett. 

“Within the decade, we will see it have quite significant transformational impacts on society. Many of those will be benefits, but some of those will be very challenging issues that we will need to face.”

These would include the possibility of taxing robots, a radical solution proposed by influential technology industry figures such as Bill Gates for dealing with the possible lack of tax revenues from the changing number of human roles required by organisations. 

We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.