The UK government is stepping up its push to build sovereign artificial intelligence capabilities, announcing a new partnership with Meta to develop open-source AI tools for transport, public safety, and defence.
The initiative, as reported by Reuters, reflects a broader effort to embed AI directly into government operations while keeping control of data, systems, and decision-making processes within the public sector.
As governments worldwide race to unlock the economic and strategic value of AI, the UK is positioning itself to reduce reliance on closed, commercial systems.
Prime Minister Keir Starmer has framed AI as a key driver of productivity across government, with the potential to modernise infrastructure and strengthen national security without outsourcing core capabilities.
Building sovereign AI capability
Under the programme, a team of AI specialists has been recruited to spend the next year developing tools that public authorities can run and adapt themselves.
The focus is on practical applications that support the maintenance of roads and transport networks, enhance public safety management, and improve decision-making linked to national security and defence.
Rather than commissioning finished products from private vendors, the government is backing in-house development of technology designed around public-sector needs.
This approach aims to ensure that AI systems reflect the operational realities of government departments and can be updated without long-term dependency on external suppliers.
Open-source approach
A defining feature of the partnership is its emphasis on open-source technology.
As per the Reuters report, the government is planning to design the tools in a way that public bodies do not have to rely on proprietary, closed-source systems.
This is intended to improve transparency and allow departments to scrutinise how AI models function and how decisions are generated.
Open-source development also gives public authorities the flexibility to adapt systems as requirements evolve.
In areas such as transport infrastructure and public safety, where conditions differ across regions and change over time, this adaptability is seen as critical to long-term effectiveness.
Expertise behind the programme
The AI team includes a data scientist from the Alan Turing Institute, alongside university researchers with experience across a range of AI disciplines.
Their expertise covers computer vision, applied machine learning for public-sector use, robotics-driven imaging, and the design of trustworthy, safety-critical AI systems.
By drawing on academic and applied research backgrounds, the programme aims to balance innovation with reliability.
Safety and accountability are expected to be central considerations, particularly for systems that may influence public safety responses or defence-related decisions.
Meta’s role and technology
Meta announced funding for the project in July last year, outlining plans for the specialists to work with open-source AI models such as Llama.
Llama is capable of processing text, video, images, and audio, making it suitable for a wide range of government use cases.
The company has said that any tools developed under the programme will be owned by the UK government.
This ownership structure allows departments to keep sensitive data in-house and tailor systems to their specific needs.
It also aligns with the government’s aim to build AI capabilities that can be audited, modified, and reused across different parts of the public sector.
The partnership highlights a growing trend among governments to collaborate with major technology firms while retaining sovereignty over critical digital infrastructure.
For the UK, the initiative represents a strategic attempt to pair private-sector funding with public control, as AI becomes increasingly central to how states manage infrastructure, security, and public services.
The post UK partners with Meta to develop sovereign AI tools for transport and defence appeared first on Invezz
