Protecting elections from deepfakes, France to use AI to help simplify public services, and more

By on 30/04/2024 | Updated on 12/06/2024

This month’s newsletter comes on the heels of our GovernmentDX event in Washington DC, which brought together federal technology leaders and global digital experts to share insight on driving innovation for high-quality public services.

You can read a summary of what was discussed at the conference, organised by Global Government Forum, here but artificial intelligence was among the key topics of conversation.

The day before the event, the global digital leaders that Global Government Forum had brought together were at the White House for a meeting on how to deliver a secure digital government experience.

This session, hosted by the Office of Management and Budget, the National Security Council and other federal agencies, focused on the critical nature of effectively delivering digital services.

“It was great to share the Biden-Harris administration’s approach to AI with so many accomplished digital leaders from around the globe, and to hear from them about their own activities,” said Ben Buchanan, White House Special Advisor for Artificial Intelligence.

“We look forward to continued international collaboration on AI.”

It is this kind of collaboration that Global Government Forum – and this AI Monitor newsletter – exists to help enable. As always, we hope you find this newsletter on how governments are using AI useful. Please share your own thoughts and projects with our editorial team.

Jack Aldane
Senior staff writer
Global Government Forum

AI and geopolitics: The threat of deepfakes in ‘the year of elections’

Global Government Forum recently attended Fortune’s first London-based Brainstorm AI conference, a gathering of industry players and experts to discuss the big questions around AI.

Deepfakes and deeper fears: In the spotlight at the event was the threat from AI deepfakes, particularly in the run-up to several major elections in 2024. Tara McGuinness, founder and director of the New Practice Lab, began by highlighting the decline of a “common conversation” around world events in an ever-fragmenting media economy.

Sir Craig Oliver, a journalist and former Downing Street director of communications under David Cameron, agreed. He added that to the extent that a common conversation exists around AI in politics currently, most of it concerns the use of deepfakes to deceive and misinform.

“The real problem is that we’re moving further and further into our different tribes, into our own filter bubbles, and AI can come along and magnify that,” he said in the session on AI’s impact on the geopolitical landscape.

How much worse can it get? McGuiness said that the only way to stop AI running roughshod over elections now and in future would be to introduce a system of “quality control”, though she stressed “We don’t [have that] at the moment”.

Julius van de Laar, a campaign and strategy consultant at Van De Laar Campaigning, said that in some elections, deepfakes have already been seen to have affected the outcome. He gave the example of last year’s election in Slovakia.

Van de Laar also highlighted a study from Switzerland that ran tests to see whether chatbots can influence public opinion. Turns out, they’re quite good at it.

“We all have the one crazy uncle that tells you that ‘The Big Lie’ exists, that the last US election was stolen,” Van de Laar said. “Well, this study found that if a chatbot is programmed with personal information about the person it is trying to convince, that chatbot is 81% more likely to change that person’s opinion than if a person tried to convince them.”

UK gives the OK to experts on AI: Another speaker on the panel was Emran Mian, who is the current director general for the UK Department for Science, Innovation and Technology. He said that the UK government had been “much more open to bringing in expertise from outside” on matters related to AI than it typically had been in the past.

“In order to move quickly, you really need political support all the way up to the top of government, and we’ve absolutely had that,” he said.

Mian also spoke about how the UK government is working internationally to build regulation around AI.

Beyond Bletchley: Reflecting on the UK’s Bletchley Park AI Safety Summit held last year (and soon to be replicated in South Korea), he stressed “the need to build international governance” around AI safety testing. At home in the UK, the priorities of the AI Safety Institute would be to build what he called a “connective tissue” between countries that have frontier AI firms, and work to understand the science of AI safety.

When pressed on the limits of international cooperation, Mian was clear that the UK would protect AI research from conflicts of interests where any might arise between itself and countries such as China.

“Our approach is to establish the basic science, what good safety testing looks like. The digital markets that we’re talking about here are markets that we already regulate in important ways,” he said.

Read more: Earlier this year, Global Government Forum published a five part report into the attempts by government-backed actors – particularly in Russia and China – to influence election outcomes and national debates in the democratic world, with the use of tools including disinformation campaigns, election hacking and party donations.

Read all five articles:

On Russia’s goals: Organised chaos: how Russia weaponised the culture wars
On Russia’s tools: Russia’s elections toolkit: dollars, disruption and disinformation
On China’s operations: A subtle opponent: China’s influence operations
On combating foreign interference via intelligence work and transparency rules: Knowing and showing: how intelligence and transparency can combat electoral interference
On improving media regulation and public education to tackle disinformation: How to create a healthier national conversation

Tech leaders to serve on US government’s AI safety board

Several major players, including OpenAI CEO Sam Altman, Microsoft chief Satya Nadella and Alphabet’s CEO Sundar Pichai are joining the US government’s Artificial Intelligence Safety and Security Board.

What’s the back story? Private enterprise has had its feet held to the fire in recent years over safety measures in AI adoption and development. Last year, the Biden administration ordered the creation of an AI safety board, following an executive order on regulating AI. The executive order specifically made “AI experts from the private sector and government” central to “advis[ing] the secretary and the critical infrastructure community”.

Board role: The AI board will collaborate with and advise the US Department of Homeland Security about how to embed AI within the country’s critical infrastructure without adverse effects. They will give recommendations to power grid operators about the implementation of AI components in their systems, as well as to transportation service providers and manufacturing plants about cybersecurity measures that are able to keep pace with the evolution of this technology.

Brains of the board: Along with Altman, Nadella and Pichai, the board will also consist of Nvidia CEO Jensen Huang, Northrop Grumman CEO Kathy Warden and Delta Airlines CEO Ed Bastian, as well as a range of other leaders in the tech and AI industry.

Register for training: How artificial intelligence can empower the civil service (30 May 2024)

France to use AI to help simplify public services

In a bid to reduce bureaucratic complexities, the French state is looking to improve and simplify the way citizens access its administrative services. As well as extending drop-in centres to 300 “medium-sized towns” by 2026, taking the total number to 3,000, and simplifying language on forms and websites, AI will also play a role.

AI – made in France: Last week, prime minister Gabriel Attal announced the implementation of a French-made AI system to streamline administrative procedures.

Introducing Albert: Attal introduced “Albert,” an AI tool designed to assist public officials in answering frequently asked questions, to save time and improve efficiency. The tool will be used by tax agents to handle queries, for example. Each response will be validated or modified if necessary by an agent, the prime minister said.

Less painful, more interesting: “The analysis of regulations will be automated, responses drastically accelerated and the work of agents made less painful and more interesting,” Attal said.

There are also plans to use AI to speed up environmental project submissions for wind farms and urban developments and to automate tasks such as the transcription of legal hearings, complaint filings and medical reports.

As elections loom: The government also announced its aim to give people the ability to delegate their voting rights via a fully digital process for all future local and national elections, starting in June this year. Since mid-April, a new version of the French identity card has made it possible for a French citizen to give their proxy for the European elections online, without validation from a police station or gendarmerie.

This method is also expected to be available in the municipal elections in 2026, as well as the presidential election due in 2027.

IRS commissioner hints at AI’s role in future tax collection

The commissioner of the US Internal Revenue Service (IRS), Danny Werfel, recently outlined two ways in which the agency is seeking to use AI technology to improve its services.

Werfel has said the agency would explore the use of virtual chatbots to assist tax-paying citizens, as well as AI tools to identify potential tax frauds in its system.

“Right now, I believe that there are AI solutions that we have not yet leveraged that exist today that can help with some of these basic questions to the benefit of taxpayers,” Werfel said.

“And on the other side of the equation, we are using AI today to do even more to unlock and spot this complexity.”

Chessboard tactics: Werfel likened the IRS’s operations to a game of chess. The task, he said, is to spot how money “is moved across different subsidiaries”. He made a distinction between “legal” and “illegal” moves, which he said AI could help governments keep track of. He stressed, however, that any decisions taken based on this information would ultimately need to be made by a human being.

As he put it: “We want the computer essentially helping us be a better chess player.”

Treading carefully: The agency will use IRA (Individual Retirement Arrangement) funds to carry out its experiments with chatbots and other solutions. Werfel said that no solution would be rolled out before privacy and security issues were fully weighed and considered, adding that the IRS would stay “1,000 miles away from even the perception that AI is in any way violating the responsibility we have to protect taxpayer privacy”.

This was important to stress, not least because the agency was called out last year in a report that showed Black taxpayers were three to five times as likely as other citizens to be audited.

The smoking algorithm: Researchers behind this report concluded that computer algorithms used by the IRS to check for fraud were to blame for the disparity. Werfel’s remarks emphasised the agency’s “fundamental responsibility” to taxpayers, particularly its duty to “protect taxpayer rights”.

Register for training: Deploying AI in the civil service (19 June 2024)

Expect an increase in AI ministers, says UAE leader

At an Atlantic Council Front Page event in April, Omar Sultan Al Olama, the UAE minister of state for artificial intelligence, digital economy and remote work applications, shared insights on the UAE’s AI readiness strategy and his thinking about the technology more broadly.

New job role: One key point he made is that we can expect the establishment of more government AI leadership roles around the world in the coming years. The Gulf nation was the first in the world to appoint an AI minister, back in 2017.

‘New lifeblood’: “AI is going to be the new lifeblood, the new foundation for most governments—and for the private sector as well – and we need to have more focused government positions that are able to really look at the impact of the technology and are able to steer it,” Al Olama said.

Preparing for extreme weather: In mid-April, heavy thunderstorms lashed the United Arab Emirates causing flash floods and disrupting flights at Dubai airport. Speaking at the event, Al Olama is quoted as saying: “It’s something we couldn’t have prepared for, but what you’re going to see next time is that we will be better prepared to leverage technology to help forecast better and ensure that people have less of an inconvenience when these events occur.”

On regulation: The minister commented: “Most governments that have just gone and banned technologies because they didn’t understand them, or because they felt like a trade-off was too large, have backtracked. I don’t think it’s healthy to do that, especially if the decision was based on ignorance.”

Join Global Government Forum’s LinkedIn group to keep up to date with all the insight public and civil servants need to know.

About Jack Aldane

Jack is a British journalist, cartoonist and podcaster. He graduated from Heythrop College London in 2009 with a BA in philosophy, before living and working in China for three years as a freelance reporter. After training in financial journalism at City University from 2013 to 2014, Jack worked at Bloomberg and Thomson Reuters before moving into editing magazines on global trade and development finance. Shortly after editing opinion writing for UnHerd, he joined the independent think tank ResPublica, where he led a media campaign to change the health and safety requirements around asbestos in UK public buildings. As host and producer of The Booking Club podcast – a conversation series featuring prominent authors and commentators at their favourite restaurants – Jack continues to engage today’s most distinguished thinkers on the biggest problems pertaining to ideology and power in the 21st century. He joined Global Government Forum as its Senior Staff Writer and Community Co-ordinator in 2021.

Leave a Reply

Your email address will not be published. Required fields are marked *