Advancements in artificial intelligence and other emerging technologies are reshaping economic growth, geopolitical influence and societal stability. Governments that fail to seize this moment risk falling behind, with disadvantages in productivity, competitiveness and state capacity compounding over time. Those that succeed have an opportunity not only to drive innovation and prosperity, but to reimagine the state and strengthen the delivery of public services.
However, trust in government is low. Rising costs of living, failing public services and weak infrastructure have left many citizens unconvinced that governments are up to the task of governing, much less managing another major economic transition. Against this backdrop, AI has become a focal point for broader public concerns about jobs, energy, security and privacy, amplifying a broader warning that existing policy is failing to deliver public confidence.
Across many advanced economies, fear now outweighs optimism. This is the AI techlash.
In response, many policymakers are reaching for blunt, reactive measures such as data-centre moratoriums. Such measures might respond to immediate political pressures, but they do little to solve underlying policy challenges. Worse, they risk slowing innovation, fragmenting AI ecosystems, weakening competitiveness, and shifting economic and strategic advantage to countries that govern the transition more effectively.
Leaders are now forced to navigate a complex situation: embracing AI is not optional for long-term prosperity and security, but short-term political dynamics often run counter to it. History shows that missing major technological revolutions creates persistent gaps in productivity, industrial capacity, investment and skilled talent that compound over time. As AI becomes an increasingly important driver of economic growth and geopolitical power, leadership at the frontier will be necessary to protect national interests at home and abroad.
Navigating the techlash successfully will require leaders to take action across three core pillars.
Embrace the AI transition with confidence. Governments must treat AI as a strategic national priority and govern the transition proactively rather than reactively. This means setting a clear direction and demonstrating, through action, that the state is prepared to harness AI to strengthen prosperity, security and public services.
Build the institutions needed to govern in the age of AI. This is a technology that demands a more capable state, not simply new regulation. Governments should strengthen institutional capacity by building modern digital capability, robust AI-assurance ecosystems and the machinery needed to manage risk, support innovation and deliver better public services.
Deliver a reimagined state for citizens. AI’s promise remains abstract for most people. Governments should prioritise practical, high-impact applications that improve everyday public services, demonstrating tangible value while building the trust and momentum needed for broader transformation. The ultimate goal is to redesign government for the 21st century, combining AI with new capabilities and reformed institutions to deliver better outcomes.
Handled well, the AI transition can strengthen economies and restore confidence in governments’ ability to act. Handled poorly, it risks entrenching distrust and slowing innovation, which will shift power and advantage elsewhere.
As AI Accelerates, Public Confidence Is Lagging
In February, hundreds of protestors gathered in London’s tech centre for the March Against the Machines, one of the largest anti-AI protests to date. By contrast, only a year ago in 2025, the same organising group gathered a few dozen demonstrators.
The increase in protestors is a symptom of a deeper tension. AI is a foundational technology that will shape economic growth, expand geopolitical influence and transform how governments operate. Individual countries cannot stop this transition. Frontier AI development will continue regardless of national policy choices.
Yet while the strategic case for AI is becoming stronger, public confidence in some parts of the world is moving in the opposite direction. Across major Western economies, worry about AI’s impact outweighs optimism and excitement. Trust in AI is low in countries such as the US, UK, France and Germany, and polling by the Tony Blair Institute has shown that, in the UK, adults are more inclined to view AI as a risk to UK society (45 per cent) than an opportunity (16 per cent).
By contrast, trust is much higher in emerging economies where AI has been framed as an opportunity for growth and advancement, rather than a threat to jobs. For example, in India, 76 per cent of people trust AI; in Nigeria it’s 79 per cent and in China it’s 68 per cent.
In developed economies, this outcry is driven by concerns often rooted in deeper anxieties predating the technology itself. AI is amplifying existing fears about economic security, the cost of living, online safety, concentration of power and national security. Employment is perhaps the clearest example. TBI polling has shown that people in the UK are nearly twice as likely to see AI as an economic risk as they are to view it as an opportunity. Meanwhile, labour-market impacts are front of mind for more than 25 per cent of UK workers, and 70 per cent of Americans fear job losses as a result of AI.
In many Western economies, people have also linked the high energy demands of data centres with their lived experience of rising energy costs; whether or not the causal connection exists, the concern is real. And then there’s online safety: AI is intensifying long-standing concerns about fraud, harassment, children’s health and other related factors, many of which significantly predate the widespread availability of AI.
These concerns are compounded by unease about the growing concentration of AI capability in the hands of a small number of firms, as well as uncertainty over whether governments have the capacity to govern the technology effectively. In turn, this reinforces a broader perception that governments are struggling to keep pace with technological change and, in doing so, becoming less capable of delivering for citizens in the age of AI.
Faced with growing pressure to demonstrate that they are responding to legitimate public concerns, many governments have reached for visible but reactive interventions. While decisive action is sometimes necessary, these measures too often respond to the symptoms of the AI transition rather than address its underlying challenges. The result is policy that tends towards blunt restrictions, reinforces public anxieties and risks closing off critical pathways for innovation. Efforts to pause or ban the development of data-centre construction demonstrate this clearly: for example, misgivings about energy capacity and how much it costs led to temporary grid constraints around Dublin, attempting to pause new data-centre connections for several years. Meanwhile, legislators in the US have proposed a national moratorium on new AI data centres, and the number of local jurisdictions passing restrictions continues to grow.
Though such steps may appear to address immediate concerns, the long-term consequences of allowing these approaches to become the dominant policy response should not be underestimated. A clear example is the public and political backlash that followed nuclear incidents in Chernobyl, Fukushima and Three Mile Island. While these were all serious accidents, public anxiety – reinforced by reactive political decisions – contributed to decades of stagnation in nuclear energy; this delayed innovation, limited deployment and denied many countries the opportunity for cheap, clean and plentiful energy. The consequences are still felt today through higher emissions, less resilient energy systems and lost economic opportunity.
AI has not had an equivalent moment. But if governments continue to prefer blunt restrictions over better governance, they risk repeating the same mistakes: slowing the development of a strategically important technology without addressing the concerns that gave rise to the backlash.
The challenge for leaders is therefore not to choose between embracing AI and restraining it. It is to govern the transition well: addressing legitimate risks while ensuring that society captures the profound economic, social and strategic opportunities AI offers.
How to Reboot Public Trust
Successfully dealing with the techlash will require governments to lead societies through a period of profound change. This can only be done by addressing the issue head on; it is not possible to retreat from technological progress. Leaders must govern the transition deliberately, addressing legitimate concerns where necessary, but also ensuring that their countries are fully able to capture AI’s benefits. There are three core principles that should underpin these efforts.
1. Embrace the AI Transition With Confidence
Governments must be proactive, rather than reactive, to address the AI transition. Leaders need to develop clear and coherent strategies that demonstrate how they intend to harness the technology to drive positive outcomes for their societies, and how they intend to address legitimate risks that could accompany the rollout of increasingly capable AI-based systems. Rather than relying on knee-jerk responses such as bans and moratoriums, governments should pursue targeted “yes, and” interventions that address public concerns without unnecessarily constraining innovation. The objective is to build confidence through decisive policy, visible delivery and clear communication, thereby ensuring a successful transition to an AI-enabled economy.
Priority action areas include:
Delivering a credible labour-transition programme. Governments cannot afford to wait for labour-market disruption to become apparent before acting. While the scale, timing and distribution of AI’s knock-on effects remain uncertain, the direction of travel is clear: if the status quo is maintained in the near term, there will likely be rising unemployment (which will place greater demand on public services), increasing wealth concentration and falling income-tax revenue for governments (meaning decreased resources to support public services). Leaders should begin to develop and test policies that can support workers through the transition. Numerous economy and policy experts have suggested changes to education, workforce-skills training, and tax and wealth redistribution mechanisms that could be used to pre-empt a shock, and set society up for success. As AI adoption accelerates, governments should continue gathering evidence to refine these policies. Preparation must begin now.
Advancing clean-energy strategy and public narrative. Data centres demand significant energy and leaders should not hide or understate this. However, governments should work to reframe the tension between energy and AI through proactive policymaking, ensuring the growth of compute infrastructure supports cheaper, cleaner and more resilient energy systems over time. Rather than allowing AI-infrastructure development to persist as an indefinite strain on national resources, leaders should use it as a catalyst to create a better and faster energy system fit for a future where electricity powers a growing share of the economy.
AI is already transforming the energy landscape. Data-centre power requirements are accelerating private investment into new technologies (such as advanced grid and nuclear technologies) and new business models that can help make the rest of the energy transition faster, easier and more cost-effective. Governments should adopt a dedicated AI and energy strategy that adequately factors AI-infrastructure needs into long-term-demand plans. They should also work with AI-infrastructure providers, regulators and energy companies to develop innovative solutions, so that AI data centres can be more effectively integrated into national grids. This could include novel models such as demand-side flexibility, as well as on-site generation and storage.
Strengthening trust in the online information environment. AI is accelerating concerns about online safety, children’s wellbeing and information integrity, but governments should resist broad restrictions that risk undermining freedom of expression. Instead they should focus on clearly defined, high-risk contexts within which intervention is both proportionate and justified. This could include efforts to protect children online, combat the erosion of the online information environment and non-consensual imagery, and clamp down on AI-enabled fraud. These efforts could be supported by the broader development and rollout of digital ID, which makes it easier to verify legitimate users, improve trust online, strengthen transparency and personal data control, and drive improvements in digital service delivery.
2. Build Institutions Fit to Govern in the Age of AI
Targeted interventions can address some of the immediate concerns surrounding AI, but they are not enough on their own. To successfully navigate the AI transition, governments will have to demonstrate that they have the institutional capability to govern it. The techlash reflects not only concern about the technology itself, but also uncertainty over whether governments can manage profound technological change. Building public confidence therefore requires institutions that can govern AI credibly, consistently and at pace.
Governing AI is a bureaucratic challenge as much as a regulatory one. Governments must be able to set standards, coordinate across departments, procure responsibly, monitor emerging risks, adapt policy as technologies evolve and ensure that AI adoption serves the public interest. Not every country requires a comprehensive legislative framework, but institutions capable of governing AI effectively are crucial.
Building AI-assurance ecosystems should form a central part of this institutional capability. AI assurance refers to the processes used to measure, evaluate and communicate the trustworthiness of AI systems, components and development practices. This means building the institutions, markets and professional capabilities needed to test systems, certify performance, audit risks, verify compliance and clarify accountability across the AI value chain. It also includes developing the support systems that maintain quality and adherence: accreditation bodies, certification schemes, professionalisation pathways, talent pipelines, standards and regulatory oversight, for example.
Public trust cannot rest on promises of responsible innovation alone. Citizens need confidence that the AI systems used for hiring, health care, welfare, policing, education, finance and public administration are being tested against clear expectations and subject to meaningful redress when things go wrong. Developers and deployers also need clearer rules so that safety and reliability become embedded in standard market practice. Well-functioning assurance ecosystems can reduce uncertainty, lower barriers to adoption and create a “race to the top” in which demonstrable trustworthiness becomes a source of competitive advantage.
Ultimately, building these institutions is part of reimagining the state itself, so that it is capable of governing in the age of AI. Governments that demonstrate they can oversee and deploy AI responsibly will not only strengthen public trust in the technology, but also rebuild confidence in their own ability to govern through an era of rapid technological change.
3. Reimagine the State to Deliver
AI’s long-term benefits will only be realised if citizens experience them in their everyday lives. Governments continue to invest heavily in AI infrastructure and articulate ambitious long-term strategies, but they need to demonstrate that AI is improving people’s lives in the here and now, not simply promise that it will.
Well-chosen, tightly scoped and rapidly deployable AI applications can improve citizens’ everyday interactions with government while demonstrating that public-sector innovation delivers results. Early successes will build confidence in both the technology and the state’s ability to govern it, creating the political capital needed for more ambitious transformation.
With this in mind, governments should:
Set realistic expectations. Governments should be clear about what AI can and cannot deliver in the near term. AI is not a silver bullet, particularly in public services where legacy systems, fragmented data and accountability requirements limit the pace of change. Rather than overselling long-term transformation, governments should address common frustrations by committing to practical improvements and delivering them consistently.
Prioritise a set of tightly scoped, high-impact use cases. Our report on delivering AI impact highlights the need for governments to prioritise AI applications that yield immediate, tangible improvements to the user experience, without requiring the resolution of more complex ethical or institutional challenges upfront.
These use cases should focus on improvements that have an outsized impact on how citizens experience public services. They should be narrow in scope, with well-defined functions that can be layered onto existing systems without the need for structural reform, accelerating deployment. They should also have high visibility, focused on high-frequency and citizen-facing services that ensure improvements are widely experienced and directly attributed to AI integration. And the applications should be seamless and production-ready, providing a clear improvement on the status quo. These initiatives should not be the only uses of AI in government, but they can be a clear, early demonstration of the potential of AI when it is embedded in public-service reform.
Use early wins to earn political capital for deeper reform towards a reimagined state. Narrowly scoped applications can deliver immediate value and in so doing help build the trust and political capital needed to pursue more complex and time-consuming digital transformation. Government should use tangible wins of this kind to harness institutional momentum and public confidence, pursuing the deeper integration, sustained investment, and careful navigation of ethical and operational risks that is needed to realise the full transformative potential of AI.
It is from this foundation that leaders can drive forward more ambitious change to transform government, making it better able to service its citizens. Current systems are slow, unresponsive and siloed, embedding the feeling that the state is not up to the task of governing well, let alone through a moment of technological upheaval. The solution is not simply a matter of digitalising existing structures, as layering AI onto broken systems will expose cracks in government rather than repair them.
Governments must work to reimagine the state. AI and digital technologies will sit at the centre of this transformation, but technology alone is not enough. Real change will also mean building new government processes, redesigning institutions and developing the capabilities required to govern effectively in the 21st century. If this is done successfully, governments will be able to rebuild citizens’ confidence that the state can meet the challenges of the age of AI.
How a Reimagined State Can Neutralise the Techlash
The AI techlash is not a rejection of technology itself, but a signal that governments have yet to demonstrate they can manage one of the most profound technological transitions in modern history. It reflects legitimate concerns about how AI is being developed, governed and deployed, combined with declining confidence in the state’s ability to manage change in the public interest. Successfully navigating the techlash therefore requires governments to embrace the transition with confidence, build the capability to govern in the age of AI, deliver tangible benefits that improve people’s lives and reimagine the state itself.
This will not be easy. Governing the AI transition requires difficult trade-offs, sustained political leadership and institutions capable of adapting at pace. But retreat is not a viable option: AI will continue to reshape economies, societies and geopolitics regardless of the choices made by any individual country. The challenge for leaders is therefore not whether to participate in the transition, but how to govern it well.
The countries that succeed will be those that embrace AI with confidence and build the institutions and capability needed to govern it. They will demonstrate, through delivery, that technological progress can deliver better outcomes for their citizens.