Gartner’s Dean Lacheca says government CIOs must anticipate and correctly respond to increasingly disruptive forces like artificial intelligence.
Dean Lacheca is a research director in Gartner’s CIO research and advisory group.
Within that group, he is part of a global team focused on supporting public sector CIOs and technology leaders.
‘Despite the scare-mongering, ultimately, Gartner’s position is that AI will actually be a net job creator’
– DEAN LACHECA
His focus involves topics including digital strategy, digital workplace, open data, government case management and citizen engagement.
His research also includes the adoption and potential impact of emerging technology trends, such as artificial intelligence (AI), on government.
How should governments be gearing up for AI?
We talk about the future of AI a lot with governments. It is still going to take time for AI to be leveraged by government. At the same time, we are seeing narrow use cases where AI is already coming into play in areas like natural language processing and chatbots being used by government agencies, so we are seeing traction on the ground with that from a government perspective. Government CIOs need to start with narrow use cases for AI.
We are also seeing natural language processing coming to things like unstructured text that governments have in areas like fraud detection. So, there are definitely narrow use cases for AI.
Government is more comfortable in the area of machine learning where they are passively training data to train particular chatbots to deliver a tangible outcome.
But it is a long journey. If they don’t start looking at the way that they manage data now it is going to take even longer for them to get that benefit at the end of the day.
That’s why the focus is on it now. It’s not so much changing the world today, but if you aren’t looking at the way that you manage data then you are really going to struggle down the road.
AI will be a long journey for government CIOs.
Are there any outliers on this yet, such as governments that are currently showing how it should be done?
In an AI sense we are seeing narrow use cases where governments are looking at biometrics for authentication, facial recognition for security and even for fraud detection. For example, New York state is using facial recognition to spot identity theft in the driver’s licence database,
We are seeing particular use cases come out but, at the other end of the equation, obviously you would have seen broader national focus on AI as an industry.
Government can be set up as both a practitioner and as an co-creator setting the direction for the country. India recently has released its position on wanting to be a leader in AI.
E-government was in vogue around the turn of the century, but took a long time to happen. Will the same happen with AI?
What society will be willing to accept is going to vary in a global sense and that will affect the pace of adoption. Despite the scare-mongering, ultimately, Gartner’s position is that AI will actually be a net job creator.
There will be levels of automation but there will also be a significant amount of jobs created. Most significantly for governments, there will be a significant opportunity to use AI to augment how they work. They are putting smart agents behind their people to help them to be more efficient and effective in the way they deliver their services.
So, in terms of big roll-outs of AI, that is not going to be where the most progress will be noticed. It is going to be smaller problems that can be solved in particular ways that happen to be leveraging advanced analytics and intelligence to achieve it.
It might not be the fanfare of the robots taking over the world. It will be more in terms of where we have made our people better at what they do by giving them better tools. But that will take time.
Will social acceptance of AI be an issue? Some countries, like Ireland, struggle even with the idea of a national ID card and social scoring in China is giving rise to all kinds of Black Mirror fears. What sort of society will AI unleash?
It is certainly going to be different for different societies. In Australia, we went through that whole journey around identity cards and it became an election issue.
Where it will make the most penetration will be where AI solves problems that the community wants it to solve.
It is not about technology itself, but how can government help the citizen and be transparent in the way it delivers services. If people don’t have to line up to submit passport applications or to renew driver’s licences, they are more than onboard with that.
We have already seen pushback in a global sense about not wanting AI to make decisions on behalf of people and there have been debates in the media about government efforts around AI.
I think that if AI is used in a support function where it leads to efficiency and it puts greater accountability in the hands of public servants themselves, then we will see it have more penetration globally.
How would you advise government CIOs trying to stay on top of these trends?
It is about not focusing on the technology itself. It is about focusing on the problem that you are trying to solve and, in terms of making progress on AI, it is about starting narrow.
Look at the data and the opportunity that you have within your role in government and make sure you are thinking in terms of that transparent engagement of the citizen so as to be very clear about how you use the technology, and make sure it has a good outcome.
We see it playing out very quickly as a virtual assistant because it is about saving the citizen time. For fraud and fraud detection, there will be a big focus area around AI.
Get a handle on your data, understand information management, understand the policies and the social licence that can be used with data, and then start very narrow in terms of applications.
Will history look kindly on AI?
As we move forward, AI is seen very much as an umbrella technology and there are a myriad of technologies and applications that make it up.
As we go down this journey there will be significant acceptance of different types of AI and then pushback against AI depending on the society that you operate in.
My advice to government CIOs is start with a narrow focus on things that lead to demonstrable outcomes and efficiencies that benefit the citizen.
Want stories like this and more direct to your inbox? Sign up for Tech Trends, Silicon Republic’s weekly digest of need-to-know tech news.