AI intelligence: equipping public and civil service leaders with the skills to embrace emerging technologies

Artificial intelligence, machine learning, and data technologies may still be in their infancy but there is no doubt they will loom large in the years and decades to come. Karen Day looks into the importance of building civil service capability to ensure the opportunities are taken advantage of and the pitfalls avoided – and at the countries that have introduced AI training schemes in a bid to get ahead
In March the UK government revealed it was developing its first ever digital, data and technology training programme for its 5,500 senior civil servants. The move to upskill its permanent secretaries and highest-ranking non-technical directors in data and artificial intelligence (AI) is a response to a global government problem: ensuring leaders have the capability to optimise the opportunities of fast-moving, world-changing technology that has implications for global security and economic dominance and for the services delivered to citizens.
It’s no surprise that over 60 countries now have AI and data strategies in place including Singapore, Brazil, and Indonesia, in addition to the UK. The pace of change in governments across the world in the last three years has sought to match the speed of AI adoption in our everyday lives. AI is already ubiquitous from the anodyne customer service chatbots to revolutionary cancer treatment, and to the potential to create biological weapons. AI has the potential to grant enormous power to the countries that develop and control it, with predictions that whichever nation dominates AI by 2030 could lead the global economic, security and social landscape well into the next century.
Understandably the rhetoric surrounding AI is dominated by a ‘race to the top’, currently headed by the US and China. But given AI’s potential, and national security and ethical implications, governments around the world are at the core of both regulating it and optimising its growth. The questions is: do the 60 governments that have adopted AI strategies have the people needed to implement them? And do those leaders have the capability to understand where billions in investment should be directed, while being able to spot potential ethical disasters?
Back in the UK and the government has improved its record on digital capability generally, establishing the GDS Academy and providing a basics of AI course since 2019. But the most recent and sustained criticism has focused on its senior leaders. In July last year a scathing report from independent watchdog the National Audit Office (NAO), explicitly linked the underperformance of digital transformation projects with a lack of capacity among senior civil servants. “Government must learn from past experience and better equip senior leaders if it is to improve its track record of delivering digital change,” warned NAO chief Gareth Davies.
The civil service’s chief operating officer and Cabinet Office permanent secretary, Alex Chisholm, conceded to a significant gap in expertise in its senior ranks last year. “We recognise that the top of the civil service, the approximately 5,500 members of the senior civil service, are people who all need to have digital and data skills,” he told a committee hearing. He outlined the introduction of a new data masterclass for senior officials, while a new digital and data sub-board has been established for permanent secretaries to build their capability and to help coordinate data, policy and operations.
“We need to keep pushing forward in order to develop more of that board-level capability and keep digital high on the agenda,’ Joanna Davinson, the executive director of the Central Digital and Data Office told the same hearing.
DDaT Essentials
The new skills programme that is under development comes on the heels of the changes announced by Chisholm and the publication of the UK’s 10-year AI strategy. Dubbed DDaT Essentials for senior civil servants, it is expected to embed digital skills within the senior ranks and benchmark current capabilities. The Cabinet Office says it is too early to give specifics on how the new programme will work but that it will “set out the skills expected of senior civil servants in the primary areas of data, technology, digital essentials, users and innovation”.
Gavin Freeguard, associate at think tank the Institute for Government says basic literacy in AI and data is essential for senior civil servants. “You want your leaders to be able to cope, not necessarily to code,” he says. “It’s important we have AI, but it’s easy to get carried away without a general level of AI and data literacy. It’s understanding the social questions, not just the technical ones. It’s about ensuring that all decisionmakers in the civil service have a sense of what they are dealing with.”
Rick Stock, managing director of RS Digital, a London-based digital consultancy that works with government, adds that senior leaders also need to be “smart customers” to spot opportunities as well as risks. “Leaders should have the digital competency to have an informed conversation in the same way you would expect them to talk with the CFO. You wouldn’t expect a senior civil servant not to understand those financial concepts and it needs to be the same with digital.”
Significantly, the UK government has exported its basic digital training model to Canada. In 2018 the Canadian government set up its own digital academy after a delegation from the UK spent a week in Ottawa helping to train its civil servants. Housed within the Canada School of Public Service it now offers over 180 different training courses on digital, data and design with the aim of “broadening and deepening digital literacy”. It is also running a 6-month programme of events on AI starting with the basics and ending with a look at AI and machine learning in foreign intelligence.
According to Freeguard, training and information isn’t vital just for government leaders. “Having an informed public that is able to engage with issues such as bias is very important in keeping governments on track, making sure they make the most of opportunities and mitigate risks,” he says.
Finland and the US also take to training
Over in Finland, a desire to inform its citizens about AI resulted in training for its government workers – a programme that has since been expanded across the EU. Finland was the first European country to launch an AI strategy back in 2017 with a major focus on education. Elements of AI started in 2018 as a free online training course developed by the University of Helsinki and consultancy Reaktor. They teamed up with employers with the aim of training 1% of its population in the basic concepts of AI to raise awareness of its risks and opportunities. The Finnish government soon caught on to the programme, using it to train some of its own staff and providing funding to roll it out nationally. A year later it ‘donated’ Elements of AI to the EU, and it was translated into member states’ languages. The course is now part of the training programme for civil servants across several EU countries including Sweden and Estonia. It was made free for all Europeans in 2020-21 and has been extended into this year due to the COVID-19 pandemic. “It is important that Finland and Europe continue to lead digital and technological development,” said Mika Lintila, Finnish minister of economic affairs in January. “The Elements of AI course in part contributes to achieving this goal.”
In the US, maintaining global dominance in AI is a political imperative with the Biden administration launching the National AI Initiative last year. It has also set out a US$180bn research and development investment into “technologies of the future”, including AI. Up to US$50bn of this will go into its National Science Foundation to build on existing tech programmes across government. Upon launching the investment, president Biden conceded that US spending on R&D had dropped significantly in recent years and warned that “China and other countries are eating our lunch”.
Biden has come under significant pressure to roll out a national AI training programme for its federal workers. In 2020 the National Security Commission on Artificial Intelligence (NSCAI) warned that without a mandated federal AI training programme, the US would “lose its edge”, jeopardising its national security and global leadership. “The federal government must reform how it recruits, trains, and educates its workforce, otherwise it will continue to struggle to modernise and maintain its lead in the world,” it said.
So far, the US has focused its training on federal workers who buy and manage AI technology. The bipartisan AI Training Act was passed in December mandating a national upskilling programme for the federal procurement workforce. The programme will be delivered by the Office of Management and Budget (OMB), which has a year to implement it, and is expected to cover the science, ethics and potential dangers of AI such as privacy violations. It will be updated every two years, to keep up with AI development, and is expected to cost around US$9m over five years. The OMB will have to report regularly on progress.
Significantly, the legislation was a direct result of the call from NSCAI for mandated training with two senators taking it forward. “Federal employees must be aware of the ethical implications, risks, and benefits associated with AI. This legislation will help protect our national security, help us remain competitive in the long run, and make sure AI technology is used properly,” Gary Peters, Democratic senator for Michigan said last year on launching the legislation.
What’s clear in all this is that AI, machine learning and data technologies will have a profound effect on governments and societies in the years and decades to come. And it is crucial, therefore, that civil services have skilled people in the right positions to spot both the opportunities and the hazards as we step forward into this new world. Governments including Finland, the UK, and the US are making steps to ensure their workforces are prepared. Others would be well advised to follow suit.
Canadian Government is not competitive enough to hire new talents that can implement and manage IA and machine learning. All technology related professions (software, computer engineering etc.) are poorly compensated in Canadian Public Service. The most talented will not take a pay cut to work in government and this become a real problem.