AI ‘makes slow-moving governments vulnerable’, EU breaks new ground on AI legislation, and more

Image: Marc-Olivier Jodoin on Unsplash

Welcome to this month’s AI Monitor newsletter. In this edition, we share AI insights from the Chandler Institute of Governance’s latest Good Government Index, details of a new roadmap from the US Bipartisan Senate AI Working Group, and much more.

Thanks for reading and please get in touch to share your thoughts and news on AI developments in government.

Jack Aldane
Senior staff writer
Global Government Forum

In this edition:

• Move slowly at your peril, governments warned on AI
• EU breaks new ground on AI legislation
• US Senate AI roadmap calls for annual US$32bn investment
• Upcoming AI webinars
• Governments navigate the double-edged sword of AI
• Government urged to ‘embrace AI or miss the most exciting time to govern’
• ICYMI: The inside story of the UK Digital Academy
• Training: get the AI skills you need 

Move slowly at your peril, governments warned on AI

Global Government Forum attended the recent launch of the Chandler Institute of Governance’s latest Good Government Index, where AI was raised as one of various factors shown to be driving a wedge between the highest and lowest-performing nation states worldwide.

Move slowly at your peril: The report said that 2023 was a “watershed” year for AI, in which governments came to realise the transformative potential and unprecedented risks of emerging AI technologies. It states that in this space, “slow-moving governments are especially vulnerable”.

Speaking on a panel at the Royal United Services Institute in Whitehall was Australia’s former prime minister Julia Gillard; Dame Margaret Hodge, the Labour MP for Barking; and Kenneth Sim, managing director of strategy and research at the Chandler Institute.

Bigger than governments: Gillard said that the task of getting AI right is one that goes “beyond national boundaries” and so beyond the work of national governments. She said that AI, like climate change, presents challenges that governments can only tackle by cooperative means.

The DeepMind model: Gillard used the example of Google’s DeepMind platform, which was released free of charge to the global scientific community, to illustrate the kind of “architecture” needed for governments to collectively harness AI safely and responsibly.

“I think what we’re going to need here is a global architecture that relies on a global expert panel to sort the science and perspective and then bring it to the decision-making table,” she said.

The AI knowledge gap: Gillard also pointed to another challenge, which is the gap between the knowledge senior civil servants have of AI, and that of seasoned AI experts. AI ministers, as we mentioned in our previous newsletter, are becoming more common across governments. However, Gillard said it will become increasingly hard for someone stepping into a role like this to know where to start if they do not have the background to inform their decisions around new developments.

Why AI and climate change differ: Hodge responded with her fear that an international agreement between governments on AI would be harder to reach than Gillard suggested. The difference between AI and climate change, she added, is that while all countries have an inherent “common interest” to save the planet, a powerful new technology could easily produce differences of intent.

“In trying to get an international agreement, there’s more of a competitive element around AI [than climate change] and so you may end up reaching the lowest common denominator, which may not be enough,” she said.

A means, not an end: “When I talk to my Labour colleagues [about this], AI is central to their integration. I keep saying to them: AI is a tool; it isn’t an end in itself. There is a danger…that AI becomes seen as solutions, rather than as tools that fit into strategies.”

EU breaks new ground on AI legislation

EU ministers unanimously adopted new AI legislation on 21 May, with the laws expected to take effect next month. The final approval of the EU Artificial Intelligence Act came after a political deal reached in December 2023.

What does the legislation target? The legislation forms what has been described as a “harmonious” set of rules by which to govern the use of AI. This includes governments’ use of AI in biometric surveillance and the regulation of generative AI systems such as ChatGPT.

Risk and rights: The legislation follows a ‘risk-based’ approach, meaning the higher the risk to society, the stricter the rules. It also aims to protect the rights of EU citizens. It will apply to areas within EU law, though does provide exemptions for military and defence uses, as well as for research.

Different approaches: The ministers’ consensus establishes a legislative benchmark for the technology and is reported to have achieved more comprehensive results than its equivalent in the US, which remains based on ‘voluntary compliance’ by providers.

US Senate AI roadmap calls for annual US$32bn investment

The Bipartisan Senate AI Working Group in the US released its AI Framework earlier this month, calling for an annual investment of US$32bn to kickstart innovation. The framework outlined the senate’s perspectives on how to navigate the rapid advancement of AI innovation across the country.

10 focuses: The framework has 10 areas of focus. They include supporting US innovation in AI; AI and the workforce; high impact uses of AI; elections and democracy; privacy and liability; transparency; explainability; intellectual property and copyright; safeguarding against AI risks; and national security.

Schumer’s emergency request: One of the leaders of the working group, senator Chuck Schumer, tweeted the announcement of the roadmap following what he called “our historic AI Insight Forums over the past year”.

Schumer added in response to reporters covering the announcement that he believed the request for emergency funding was warranted. He stressed that it would strictly be non-defence discretionary spending to support efforts at the Departments of Energy, Commerce and other agencies.

“We’re going to figure out the best way to get it done. It’s not all in the first year,” he said.

Upcoming AI webinars

How governments are using AI to become more efficient
9 July

This webinar will provide practical insights on how AI technologies including automation and machine learning can be used to drive more effective, efficient and user-focused services.

Register now

How government can make the most of analytics and AI
10 October

This webinar will bring together public servants to share insight on how government can use the data it collects to better deliver for people, and also examine how governments can use artificial intelligence to optimise digital services and experiences.

Register now

Governments navigate the double-edged sword of AI

AI could boost civil service productivity but it also presents major new challenges for government, including rapid economic change and an urgent need for regulation. At the Global Government Summit, national civil service leaders debated the opportunities and threats presented by these fast-evolving technologies.

Government’s best bet for change: Former permanent secretary of the UK Cabinet Office Alex Chisholm called AI “the single best thing we’ve got to try and improve the way government works”.

Hype-cycle survivor: AI is not like other technologies that Chisholm called “false hopes of the past”. As Chisholm put it, AI is proving “genuinely transformational”, unlocking huge benefits for citizens, public servants and taxpayers alike.

Promising on productivity: Chisholm pointed to slow public sector productivity growth in many countries to make the point that applying AI technologies could improve output by 30-45%. As average ages rise across much of the rich world, he added, this could prove especially valuable as demand for services increases and the working-age population shrinks.

A range of AI skill sets: “Everybody needs to be trained in AI, but not in the same way,” Chisholm said. Senior leaders in particular must “understand its capabilities and be convinced of them, because we have control over resources and priorities within our organisations”.

Lessons from Singapore: Sim Feng-Ji, deputy secretary of Singapore’s Smart Nation and Digital Government Office, said the city-state is treading carefully on AI. The civil service first launched a wide range of programmes to improve the workforce’s understanding of AI, to raise skill levels, and to “encourage people to experiment”. These included training, online newsletters, workshops, community-building activities, and the creation of a digital test environment.

“You need to have a culture of being able to start small, and to fail quickly – and cheaply,” he commented.

Low risk, high reward: Digital leaders sought use cases that were “relatively low-risk, but quite easy to adopt”. As examples, Sim Feng-Ji cited search functions guiding citizens to the right public services, and transcription software to aid civil service note-taking. More than 500 such projects are now up and running, he added.

Managing risk: Sim Feng-Ji’s office inspects and approves proposed applications via an AI Development Group, which includes representatives from across government. Over time, departments steadily expand their use of AI into more advanced and complex fields.

His team are particularly alert to legal risks, he explained, and to data security: AI proposals are checked carefully to ensure that providers won’t retain data, that sensitive data won’t be compromised, and that there’s no danger of de-anonymisation.

Read more: Catch up on the whole session summary

Government urged to ‘embrace AI or miss the most exciting time to govern’

If the UK government doesn’t grab AI by the horns, it could miss a “historic opportunity” to govern in an exhilarating new era, according to a paper by the Tony Blair Institute.

Get a grip: The report was published on 20 May, a couple of days before UK prime minister Rishi Sunak announced a UK general election set to take place on 4 July. It concluded that “if gripped properly”, AI could herald “the most exciting and creative time to govern”.

A public sector ‘on its knees’: “The public service [has] large backlogs and lengthy waits for services, a demoralised, unproductive workforce and a lack of long-term thinking as policymakers go from crisis to crisis,” the report said. Adopting AI in the public sector would be a question of “strategic prioritisation that supersedes everything else”.

Other voices: The institute isn’t the only large organisation banging this drum. At a presentation in April for AITalks, an event for top government leaders, tech innovators and industry experts to explore AI in the public sector, Dr William Chappell, vice president and chief technology officer of Microsoft’s Strategic Missions and Technologies Division, outlined the three waves of AI in government and what they mean for agency operations.

The three waves: The first wave Dr Chappell described involved attempts to embed human expertise in hardware. The second wave was around statistical learning and AI’s ability to recognise events and objects. The third and current wave is based on “contextual adaptation”, where AI models can broadly understand things through a combination of perception, learning, abstraction and reason. And it’s this third wave that Dr Chappell said marks a shift from AI as used in “bespoke models” to serving “as a foundational tool with myriad applications across diverse domains”.

ICYMI: The inside story of the UK Digital Academy

In the latest episode of Government Transformed, GGF’s executive editor Richard Johnstone and podcast producer Jack Aldane journeyed from London to Leeds to join a reunion of one of the UK government’s most innovative digital projects – the Digital Academy.  

Listen now: This podcast offers a timely example of the difference small groups of dynamic individuals can make to government services, given enough time and freedom to solve core problems with legacy systems. It also shows why digital technology leads to transformation only when people combine to form a mission-driven culture.

Training: get the AI skills you need  

Global Government Forum’s expert government training team create and deliver training programmes for governments all over the world. Some key upcoming courses include:
 
How Artificial Intelligence Can Empower the Civil Service
30 May  

This interactive workshop is designed to introduce civil service professionals to the world of AI. The seminar will cover fundamental concepts of AI, its applications in the public sector, ethical considerations, and practical tools. It aims to provide a comprehensive understanding of AI, especially for those new to the field, and explore its potential in enhancing government services. 

Deploying AI in the Civil Service
19 June 

This seminar is designed to provide a full understanding of how to go about using and deploying AI in civil service departments and organisations. It starts with the fundamental questions, and moves into practical considerations, legal frameworks and organisational compliance. It uses practical case studies and also looks ahead at what the future holds and how civil servants can prepare.

Thanks for reading this newsletter and keep an eye out for next month’s edition.