Recent government announcements on AI show that it is serious about using AI to grow the economy and improve public services. There is increasing hope and optimism around the potential benefits of AI for the public sector, from AI assistants to automation across education, health and policing. The Government wants to ‘mainline AI into the veins' of the UK and the Prime Minister has described AI as a ‘golden opportunity' for reforming the state.
However, this ambitious agenda will not succeed without the right approach to procurement, governance and responsible innovation. Done badly and without due consideration for its broader impacts, the rapid roll-out of AI across the public sector could lead to poor outcomes for people and society, especially marginalised communities.
At Ada, we have seen how local government has an important role to play in the responsible deployment of AI. Yet, between local government and industry, current procurement processes, governance structures, market dynamics and unequal AI expertise are all barriers which stack the odds against success.
When viewed against a backdrop of budget cuts, devolution and a currently fragmented and duplicative approach to digital technology, the Government's ambitions for improving public sector finances and delivering better services through rapid adoption, while also maintaining public trust, seem like competing and at times unattainable goals.
The public sector faces a number of data and technology challenges, however the overall response to these issues remains uncoordinated and disjointed. This creates friction and duplication which hinders efforts to realise the benefits of digital and AI within the sector.
A more cohesive and coordinated approach is urgently needed. Local government can lead the way in this through its approach to the responsible procurement and deployment of AI technologies.
However, despite a growing data bank of AI use cases and ongoing collaborations with industry, local government efforts are often overlooked in national policy.
When central government departments are engaged with local government initiatives, this is usually on an ad hoc basis, rather than the result of any systematic or comprehensive programme. They also often replicate the top-down approach that has plagued relationships between central and local government.
At Ada, we have seen how local government has an important role to play in the responsible deployment of AI. Yet, between local government and industry, current procurement processes, governance structures, market dynamics and unequal AI expertise are all barriers which stack the odds against success.
Local authorities seeking to address some of these challenges often duplicate efforts, run into siloed working models with limited access to shared knowledge or best practices - all of which frustrate those responsible for AI procurement.
From our interaction with experts, one thing is clear. Central government has the power to reshape the conditions in which local government operates. Small and medium enterprises (SMEs) are more likely to supply products to local government, however market imbalances and monopolies still allow larger suppliers to dictate the pace of change and adoption.
Local authorities on their own lack the power to reshape procurement frameworks to better suit their needs, which leaves them reliant on suppliers to uphold the same values and high standards demanded of public bodies.
This is why we're calling for a national taskforce to reform the procurement of AI in local government. With the right membership, governance and powers, a taskforce could help ensure local government has the right tools and frameworks necessary to buy and use AI responsibly.
Despite the many policy announcements over the last few months, including new forms of engagement between central and local government, the need for a taskforce has only grown stronger.
In its recent report the Public Accounts Committee underscored the immense scale of the challenge: ‘The Government has said it wants to mainline AI into the veins of the nation, but our report raises questions over whether the public sector is ready for such a procedure.'
A taskforce could take the lead on pushing forward many of the Committee's recommendations. For example, by independently collating evidence on AI pilots and deployments, the taskforce could guide future investments and prevent wasted resources - not just in local government but across the wider public sector.
Our view is that a national taskforce should prioritise:
• Mapping different uses of AI in local and central government to help ensure alignment between their goals
• Conducting research into procurement challenges, particularly in relation to dominant suppliers, and working with industry and regulators to address those, for example, through responsive contract templates.
• Coordinating AI training resources to provide a more consistent approach to skills development across local government.
• Establishing independent advisory contacts to support lower-resourced councils with access to knowledge, information and advice about AI adoption.
There is a lot that public sector can learn from initiatives already emerging across local government. However, as real and lasting change can only come from a taskforce uniting frontline experience and expertise with those who have the power to reshape the conditions in which councils operate.
Mavis Machirori is a senior researcher at the Ada Lovelace Institute
