The real AI enablers in transport: trust, talent, and data

The publication of the Transport Artificial Intelligence Action Plan: Transforming Ambitions by the Department for Transport (DfT), UK sets out an ambitious vision: “Responsible AI embedded in a resilient transport system delivering cheaper, cleaner, and safer journeys for all”. The AI Action Plan’s five key areas provide a comprehensive framework for transformation, but success hinges on foundational enablers that determine whether AI systems gain acceptance, deliver value, and scale effectively.

Drawing from our collaboration with the DfT and other major transport organisations across modes, the real enablers of AI success aren’t just technological, they’re human and institutional. Trust, talent, and data will be critical factors in turning the vision of better journeys into reality, and the Government has a key role to play in this process.

Trust is fundamental to AI adoption

Public support for AI in transport depends on ethical standards, explainability, user involvement, and trust. The opportunity is significant: when trust is established through engaging with the public and demonstrating practical, useful AI applications, adoption accelerates rapidly and delivers transformational benefits.

Building trust requires establishing clear ethical standards that guide AI development and deployment, ensuring systems can explain their reasoning in terms users understand, particularly for safety-critical decisions. User involvement throughout the development process will create advocates rather than sceptics, whilst demonstrating reliability and safety through real-world applications will build confidence over time. This means moving beyond technical excellence to actively engage with communities, address concerns proactively, and show how AI systems improve transport services in ways people can see and experience directly.

We supported Eurostar to build and drive adoption of an AI-driven prediction model, showing how trust is fundamental to adopting new technology. By transparently involving operational teams to test and refine the AI solution, and using only existing data without invasive new collection, Eurostar addressed concerns around change and privacy. Co-development sessions empowered staff with the understanding and skills to use the AI system, which helped ensure buy-in. This trust-based approach is enabling Eurostar to confidently deploy AI for station congestion management and improved passenger experiences.

Skills will make or break AI delivery

Investment in AI fluency across the transport workforce is essential, including day-to-day productivity and development of AI, digital, and data skills. Demonstrating practical, useful AI applications to people, rather than just discussing the theory, is crucial for building the necessary skills and adoption across the transport sector.

The AI skills challenge in transport isn’t just about hiring data scientists, it’s about creating organisation-wide capability to work effectively with AI-powered systems. This requires a fundamental shift in how we think about professional development, moving beyond abstract training to create environments where transport teams experiment with AI tools using real data and scenarios. The AI Action Plan’s emphasis on internal AI training, and creating a Transport Sector AI Community of Practice, recognises this collaborative approach, ensuring learning is shared across organisations whilst developing practical capabilities for the future. Success comes from building confidence alongside competence, enabling staff to understand not just how AI works, but when and how to apply it effectively within their specific context.

We have worked with the DfT to design and rapidly deploy targeted AI training pathways tailored to different staff roles, ensuring content is relevant and engaging to drive adoption and compliance. This is delivered on a Learning Experience Platform to help DfT to embed AI skills at scale.

There are lessons the transport sector can learn from other industries on how to embed AI skills, particularly with frontline workers. Our work with West Midlands Ambulance Service (WMAS) included supporting more than 20 frontline workers to develop their own low-code applications using Microsoft Power Platform. These ‘citizen developers’ were supported by the Trust’s digital team to put in place the processes and governance that allow the best apps to be used by the relevant WMAS teams.

Data as shared infrastructure is the foundation

Fragmented, poor-quality data remains a significant barrier to AI adoption across the transport sector. The UK’s Transport Data Strategy aims to set the conditions for productive use of transport data, focusing on improving data quality, establishing shared standards for data sharing and interoperability, and addressing data protection concerns.

The fundamental challenge isn’t just technical, it’s organisational. Transport data exists across multiple operators, modes, and jurisdictions, often in incompatible formats with inconsistent quality standards, creating a coordination problem that individual organisations cannot solve alone. What makes this particularly complex in transport is the need to balance commercial sensitivity with collective benefit. Unlike other sectors where data sharing is primarily internal, transport requires unprecedented collaboration between competitors, regulators, and public bodies. The Transport Data Strategy recognises this unique challenge, establishing frameworks that enable data sharing whilst protecting commercial and personal interests. Success requires treating data not as a byproduct of operations but as shared infrastructure, implementing automated quality assurance processes, creating governance models that enable productive collaboration, and establish clear standards that work across different systems, maintaining competitive advantage where appropriate.

We are developing the Digital Traffic Regulation Orders (DTRO) data platform for the DfT and local authorities with the core belief that ‘data is the foundation’ for enabling future AI innovation for safer roads. By creating an API-first, digital platform that standardises TRO data to make it openly accessible, we are ensuring that high-quality roads data is available for a wide range of transport use cases. User-centred design and comprehensive validation rules increases data quality and modernises regulatory processes, while laying the groundwork for future AI applications in transport.

Government’s enabling role is essential

To accelerate safe and scalable AI adoption in transport, the Government should actively champion a leadership model focused on standards, capability-building, and broad engagement across the sector. This means proactively driving collaboration between public bodies, local authorities, industry partners, and communities to unlock AI opportunities and address barriers such as skills gaps, public trust, and fragmented infrastructure.

The Government is uniquely positioned to invest in collective enablers, such as common standards, shared infrastructure, and open frameworks. These will benefit the entire transport ecosystem, rather than prioritising narrow competitive advantage. By offering regulatory clarity and adaptable governance, the Government can unlock greater private investment while preserving flexibility for innovation.

Crucially, effective AI governance must go beyond technical solutions to foster social license and trust. Government should facilitate knowledge sharing, coordinate investments in foundational infrastructure, and ensure that AI development aligns with broader transport system objectives. Rather than relying solely on regulation, the Government can play more of an active orchestrating role bringing together diverse stakeholders, incentivising collective action, and driving capabilities that no single organisation could develop in isolation. This ecosystem approach is vital for establishing a robust, future-ready environment for responsible AI adoption in transport.

The DfT AI Hackathon in Transport, delivered in partnership with PA and Google, is a clear illustration of how the Government’s enabling role is essential. By bringing together leading industry players, transport authorities, and top academic institutions, DfT created a unique space for cross-sector collaboration and rapid ideation on pressing transport challenges. The hackathon enabled participants to share data, experiment with AI solutions for issues such as congestion and accessibility, and build practical prototypes in a supportive, government-led environment. This targeted, government-facilitated approach proved instrumental in breaking down silos and accelerating innovation that addresses real-world needs in transport.

Delivering the step change to embed AI in transport

The Transport AI Action Plan provides a comprehensive roadmap for transformation, recognising that “AI is already delivering real impact across the transport system, but we are now on the cusp of a step-change in opportunity.” The specific actions from developing leadership statements for AI implementation to creating innovation partnerships acknowledge that success requires systematic attention to trust, skills, data quality, and coordinated leadership.

Organisations that invest in these enablers now will lead the transformation of UK transport. They will build trust through demonstrable safety and transparency and develop AI fluency that bridges domain expertise with technical capability. Organisations will also be able to create the data architecture that supports AI at scale, and actively engage with the collaborative transport ecosystem.

The opportunity is unprecedented, the framework is clear, and success depends on executing the fundamentals that enable responsible AI to deliver cheaper, cleaner, and safer journeys for all.