Recognising the persistence of the issue, the Government has renewed its focus through the Digital Inclusion Action Plan: First Steps, published in 2025, which outlines a more coordinated approach across departments and stakeholders. Its initiatives include a Digital Inclusion Innovation Fund to support local initiatives and scale approaches that work, a strengthened Essential Digital Skills framework, a pilot device donation scheme across government departments, improved usability of digital government services through GOV.UK One Login, and a stronger evidence base on what works and the economic case for investment.
The Action Plan explicitly acknowledges that digital exclusion cannot be solved by the central government alone. It requires partnership with industry, local government, charities and community networks. Voluntary organisations and social enterprises continue to deliver frontline digital skills support, while banks, telecoms providers and technology companies have launched affordability schemes and training programmes.
More recently, the UK has also expanded its artificial intelligence (AI) skills initiatives, including the recently launched AI Skills Boost, a government-industry partnership that aims to improve UK workforce readiness by upskilling 10 million workers in AI skills by 2030, with training courses designed to help people get comfortable with new technology and increase AI uptake. This ecosystem of effort is significant, yet exclusion persists, because the gap is structural as much as it is technological.
AI cuts both ways
AI has brought both opportunities and risks when it comes to exclusion. On one hand, it could widen the divide. As services become AI-enabled, a new layer of capability is required to understand how AI works, how to question outputs, how to identify errors, and how to use tools safely. This AI literacy gap risks layering a new inequality on top of existing digital skills gaps. And if AI systems are poorly designed, biased, opaque or mandatory without alternatives, they could increase confusion and distrust, particularly among those already marginalised.
On the other hand, AI also has powerful inclusion potential. Used responsibly, it can improve accessibility through speech-to-text, translation and adaptive interfaces, provide step-by-step guidance for complex processes, personalise learning and skills development, and support frontline workers to deliver more tailored assistance.
AI could reduce barriers, but only if inclusion is embedded by design. Without intentional action, it is more likely to accelerate existing disparities.
What needs to happen next
To ensure AI becomes an inclusion accelerator rather than an exclusion amplifier, both government and industry must take deliberate action in five areas.
1. Treat AI literacy as foundational
AI understanding cannot remain a specialist skill. It must be integrated into essential digital skills frameworks, workplace training and community learning, especially targeting those with low confidence or low baseline skills. The AI Skills Boost programme is a start, but it needs to be activated across the groups who are experiencing exclusion.
2. Invest in the fundamentals
Devices, connectivity and local support networks remain critical. AI cannot bridge a divide where basic access is missing, and sustainable funding for community-based digital skills support is essential.
3. Embed participatory service design
Perhaps most importantly, digital and AI-enabled services must be designed with, not just for, the people who use them. That means co-designing services with older people, disabled people, carers, low-income households and minority communities; testing services on older devices, low bandwidth and shared access conditions; providing assisted digital and non-digital alternatives; and building feedback loops so services evolve based on real user experience. Inclusion cannot be retrofitted after deployment. It must be a starting principle, and designing for the most excluded will benefit everyone.
4. Prioritise trust, transparency and accountability
One of the most needed changes, and one of the most difficult to deliver, is trust. We need to address issues such as bias, hallucinations and malicious misuse transparently, and ensure we understand where appropriate human intervention is required, so that AI-enabled services can be trusted. That includes clear explanations, redress mechanisms and human escalation routes. Trust is a precondition for adoption, and without it, even well-designed systems will be underused.
5. Make this a shared accountability agenda
Digital inclusion should be treated as core infrastructure. It underpins economic growth, productivity and public service reform, and that requires alignment between policy, industry investment, skills systems and service design standards.
The UK has strong foundations – an established Essential Digital Skills framework, a renewed government Action Plan, active civil society networks, and growing AI capability. But coordination and execution are critical. Improving digital inclusion requires strategic alignment among digital, data, and inclusion agendas; robust data and insights to target interventions effectively; responsible AI governance embedded in transformation programmes; participatory service design methodologies; and cross-sector partnership models that move beyond pilot funding.
As AI reshapes how services operate and how citizens engage, digital inclusion must move from the margins to the centre of transformation strategy. Technology will continue to transform society, but it is up to us whether that transformation expands participation or widens the exclusion gap.
With the right combination of policy, industry leadership, community delivery and inclusive design, we could bring people together to embrace the future. If we fail to act intentionally at this point, we could simply create more division, more social inequality and a gap that is too difficult to close.
Gita Singham-Willis is strategic engagement director at Transform
