Head of Digital Services for Government and Transport, BAE Systems Applied Intelligence
20 Dec 2018
Artificial Intelligence (AI) offers huge potential to governments but its positive impact is by no means guaranteed. Harjit Lota explores how to turn barrier into breakthrough.
I had to renew my passport the other day. Not so long ago I would have had to take a trip down a bureaucratic lane. But this time it took just 10 minutes online from my sofa. This is just one example of how government services are being transformed by digital technologies.
Tapping into technology
Businesses across a range of sectors are increasingly recognising the benefits of AI. At BAE Systems, we have been at the forefront of developing AI technology for many years, and not just in the defence sector.
Today, policymakers are increasingly turning to AI through things like Machine Learning and Voice-Assistants. AI and Robotic Process Automation (RPA) are now seen as enablers of innovation across government, according to the 2018 technology innovation in government survey. These technologies are expected to feature in the 2019 innovation strategy for government – being led by the Government Digital Service.
These technologies can be used at the front-line of public services, as well as back office functions crucial to a smooth-running government. Helping make predictions from data, improving productivity by optimising and speeding up operations, and automating and improving the accuracy of software transactions are just a few of the ways they can have a real impact – and this is already underway.
HM Revenue and Customs (HMRC) and the Department for Work and Pensions are leading the way. More than 12,000 software robots are being used to automate transactional IT processes. In HMRC’s case, they are helping 7,500 contact centre advisors automatically open up case files, reducing 66 mouse-clicks down to 10 per call.
New tech, same old barriers
This is all greatly encouraging but there remains much still to do. From our experience of helping with technology-based change over the past 40 years, there are some important barriers to overcome to aid successful adoption of AI technologies.
Most of these barriers are not about the technology.
In addition to the natural risk aversion and fear from not understanding AI, it isn’t clear how new technologies like AI will fit into annual budgeting and prioritisation. Investing early in AI could, in the long-term, improve how critical services such as cyber security monitoring and managing internal case-loads are delivered.
Investment includes ensuring the right AI skills are available to use these emerging technologies. As we know, the demand for data scientists and data engineers in government is hardly shrinking.
Then there are well-known issues around security and data protection. Throw in uncertainty over trust and ethics, as my colleague Holly Armitage explains together with a fear that AI will lead to greater job losses, and you’ve quite the proverbial brick wall ahead.
So, how will government seize the opportunities and benefits offered by AI and RPA?
Team up to build up
Experience tells us that government departments need to adopt a holistic mind-set by focussing on a capability-based approach across people, technology, processes and infrastructure:
Mandate. It starts at the top. From ministers through to senior civil servants, there must be a clear mandate and approval for AI adoption. The positive case for change should be clear and agreed.
Understanding. Of equal importance is helping the workforce understand these technologies and how they can help, rather than an excuse for redundancy. Ongoing dialogue with key civil service unions should be a part of this.
Re-design the business. Departments were hardly designed for the digital age. As technology changes, so should supporting business structures. Focussing on governance and policy that work with AI will be key – especially around trust and ethics for decision making.
Start small. Finally, the temptation to run before walking should be avoided. An incremental and iterative approach (e.g. proofs of concept) and tailoring to suit departmental maturity will help. This could help avoid repeats of (non-AI) IT failures that adorn much of Whitehall’s recent history.
Government should be applauded for recognising the potential of AI to improve internal and external services. But as technology advances, so should turning barriers into enablers. This is no time to press pause – quite the contrary.
About the author
Harjit Lota is the Head of Digital Services for Government and Transport at BAE Systems Applied Intelligence.