AI-pocalypse? No.
Why we need to focus on how tech is transforming Britain today
Emily Middleton and Kirsty Innes
Rishi Sunak is today meeting an elite group of technology companies and international leaders behind closed doors at Bletchley Park. Together, they will discuss the long-term, existential risks of artificial intelligence (AI). As they gaze into that distant horizon, they will overlook something far more important and immediate: the impact that technology is already having in Britain today. Rather than pontificate about the years ahead, the UK urgently needs an approach to technology that addresses the hopes and fears of working families today - as new polling from Labour Together shows.
The public are optimistic about technology
Overall, the British public is positive about the opportunities that technology can provide. Over half (53%) see “more opportunity than risk” when they think of new technologies, and just one in ten (10%) see “more risk than opportunity”. Unsurprisingly, young people are most likely to be optimistic. But even amongst over-65s, almost half (49%) still see more opportunity than risk, and just 11% think the reverse.
But voters are wary of Artificial Intelligence (AI)
When you ask people about AI, instead of technology more broadly, their attitudes flip. Now, just one in ten (11%) are “more excited than concerned”, and nearly half (46%) are “more concerned than excited”. The older the voter, the more concerned they become.
It’s not an AI-pocalypse we fear most
Rishi Sunak has brought together his elite group to imagine how AI could destroy human life, summoning nightmares worthy of science fiction. The approach might summon some eye-catching headlines, but it avoids what really concerns voters.
Our polling shows that the British public is more worried about existing threats than long-term risks. While “existential risk” is their fourth highest concern, it sits behind three far more immediate concerns. First, how AI could spread misinformation (a concern for 82% of voters). Second, the fear of job losses (73%). And third, the fear of AI being used to monitor or control people at work (72%).
The Government has acknowledged these risks in the past. Its AI White Paper, published earlier this year, did so. But having summoned leaders from across the world to tackle this genuinely difficult challenge, they have kept their blinkers on. As one tech company representative told Wired magazine last week, Sunak’s approach to the summit is “government by photo op.”
The summit is also failing to address the inequalities that are likely to result as AI’s impact on the labour market gathers pace. A recent study from the Institute for the Future of Work and Imperial College London found that the effects on jobs are likely to be uneven across different parts of the UK. The report warned, “If handled poorly… automation's transformation of the labour market could exacerbate existing regional and demographic inequalities and could lead to serious erosions in job quality, with all of the other negative impacts that that will bring”.
This problem isn’t unique to the UK
Other governments are already innovating to tackle similar issues head-on, sometimes in much more challenging circumstances. When war broke out in Ukraine, millions of people were displaced, and unemployment rose sharply in the worst-affected regions. The Government of Ukraine had an existing service, Diia.Education, which had already boosted the digital literacy of 1.4 million Ukrainians before the war. The Government’s digital unit rapidly pivoted this service to help citizens retrain in professions as diverse as video editing, online marketing, floristry, and sales.
Singapore also has extensive skills provision for adults. Launched in 2017, the government offers a two-day programme focused on digital technologies and their impacts. Participants can then undertake further training to access jobs in advanced manufacturing, social care, data science, the green economy, and other growth areas. The government’s digital agency has also partnered with technology companies on its Digital Academy to provide upskilling and retraining opportunities in digital, data and tech for thousands of public servants.
The UK could do more to leverage the private sector’s capabilities. For example, Google has trained more than 1 million people in digital skills since 2015, including more than 600,000 in small businesses through its Digital Garage programme. Courses include design, data & analytics, selling online, cloud computing, and artificial intelligence. Barclays’ Digital Wings programme improves foundational digital confidence and awareness for adults, and has helped more than 455,000 young people to start learning to code.
Voters will embrace tech if their public services improve
There is a strong appetite in the UK for new digital services that improve the provision of public services. There is particular interest among the UK public in new digital services related to healthcare and education, such as secure sharing of medical records (80% thought this would be beneficial), and online, personalised learning resources for children to use at home (71%). These services are already available in some healthcare providers and schools, but they are not universal nor nationwide. The availability and quality of services like these remains a postcode lottery.
The Shadow Secretary for Health & Social Care, Wes Streeting, has already announced his intention to enable patients to self-refer, in order to speed up access to specialists. Our polling suggests that the ability to do this digitally would be popular (68% thought this would be beneficial).
Of course, surveys are not the best way to test new digital product ideas. And digital services are not right for everyone. A third of adults we polled value being able to speak to a real human to solve their problem more than convenience (see table below). Yet better digital service provision can free up time for public sector workers to spend with those who need it. That’s because effective digital services tend to reduce failure demand. Failure demand happens when users are struggling to use a service, and often manifests as increased queries and complaints to contact centres. Allowing more people who want to, to self-serve online, would free up government contact centre staff or their equivalents.
Beyond specific sectors, there is an appetite for more seamless interactions with government. Most (64%) voters believe a “Tell Us Once” commitment would be beneficial. Pioneered by the Government of Estonia, “Tell Us Once” is the idea that the government will never ask for the same piece of information twice, thanks to secure, efficient data exchange between government departments.
The UK has experimented with this and has succeeded in some areas, such as the service to register a death, which allows users to “report a death to most government organisations in one go”. It is now used by hundreds of thousands of people each year, and also reduces the administrative burden for public servants. However, the behind-the-scenes effort in making such a service happen is phenomenal because so much coordination is needed to make it happen. As a result such services are still rare in the UK. Part of the problem is that it remains very difficult to make the case for funding services and platforms when the benefits accrue to multiple departments. Stronger ministerial support, improved data-sharing culture and infrastructure, and clearer incentives for collaboration across Whitehall, are also needed.
AI appeals if public services can be improved
We have already seen that while technology is popular, AI is a cause of concern. However, when you descend from the general to the particular, people’s attitudes to AI begin to change. Despite concern about AI technologies as a whole, the public is remarkably positive about specific uses of AI to improve public services.
Over two thirds (69%) of voters support using AI to help hospitals predict admissions and manage their beds. Identifying tax avoiders was also particularly popular (68%). So were opportunities for the Department for Work and Pensions to use AI to uncover fraud (65%). In every possible specific application of AI we tested, usefulness always outweighed concern.
Pressing need for progress, especially in health
The public’s interest in using more digital technologies in the NHS is unsurprising because the scale of the problem is so great. Each year, around 15.4 million appointments are missed, resulting in the loss of more than £216 million of taxpayers’ money. These are simple problems that cry out for better use of technology. In a recent study in one NHS Trust, the primary reason that patients were missing their appointments was that they did not know they had one. Earlier this year, the Newcastle Hospitals Trust admitted that 24,000 letters weren’t sent due to a ‘computer error’, resulting in missed appointments, crucial tests, diagnoses and results. A senior official from the Care Quality Commission flagged the pressing risk to patient safety posed by such IT failings.
In June 2023, the House of Commons Health and Social Care Committee filed a report that asserted digital transformation across the National Health Service remains “slow and uneven” with “substantial variation between organisations”. Given the scale of the challenges the NHS faces, there is a pressing need for progress.
We need to talk about trust
Most AI tools are built on access to large, high-quality datasets. To make the most of AI to improve public services, it will be essential for government bodies to earn the public’s trust, and show that they will collect, store and share data responsibly.
The opportunity and the need for technology – including AI – to help improve our public services is clear. But if the next government is to be successful in realising this potential, it will have to improve trust in the public sector’s use of personal data.
Almost two thirds of people we polled (64%) trust the NHS to look after their data. But the NHS is the only public institution that was more trusted than distrusted. The proportion of people who trust central and local government with their data is 25 percentage points lower at 39%. Over half (53% in both cases) said they had little to no trust.
The fears that underlie this distrust are broad. More than two thirds (71%) are worried their information might be inadvertently published or inappropriately accessed. A similar proportion (69%) are worried about being exposed to the risk of identity fraud. Many are also worried about their data being inappropriately shared with a private sector company, leading to targeted adverts online (69%). A majority (58%) are also concerned about their data being inappropriately shared between different government departments without their knowledge.
Given high profile data scandals, such as in the Northern Ireland Police Service, as well as new evidence of AI bias and misuse in the public sector, these results are unsurprising. While there are pockets of good practice, there is a huge amount of work to be done to build trustworthiness across the public sector to ensure data is used responsibly and effectively.
Trust and capability also needs to be built within and between different parts of the public sector. The Department for Levelling Up’s own research, published earlier this year, showed that a lack of “trust and confidence” between local and central government inhibits data-sharing. The same report also showed that local councils typically do not think they have the resources nor the expertise to properly clean data and share it securely with central government. Trust in public servants tends to follow them being worthy of that trust. On that score, there is still more to be done.
What’s next
These are knotty problems. They raise deep questions about how we resource, empower and guide our public sector. These problems won’t be solved by a select group closeted away in a stately home over 48 hours this week. Instead, they will be solved by engaging in a serious, pragmatic public dialogue that brings together diverse voices and interests. They will be solved by taking a frank look at the impact that AI is already having, and by tackling the concerns of the public head on.
Only by addressing these legitimate worries can leaders seize the opportunities presented by AI and other fast-developing technologies. We can renew our public services for generations to come. But careful work that rebuilds trust and confidence in the government's ability to handle data must serve as the foundation.
Over the coming weeks and months, Labour Together will be exploring how. We will be talking to technologists, entrepreneurs, public services experts, citizens and activists. Together, we will explore how a Labour government could:
Put in place the right teams, with the right skills and levers, to drive ambitious digital transformation throughout the public sector
Build public trust and confidence in its ability to use and share data responsibly
Help ordinary working people to gain the digital skills needed to thrive in today’s economy
Enable our towns and regions to capitalise on the economic benefits of technology
Rather than obsess about the dystopias of tomorrow, these are big questions that must be answered today. If you happen to be working on any of them, and want to help Labour Together as we do too, please get in touch!
