MHCLG names AI priority projects for English councils

TL;DR:

• The Ministry of Housing, Communities and Local Government’s Local AI team has set its priority projects after AI ethics drop-in sessions with 60 councils.
• Projects include an AI playbook for local government, standard assessment templates and expanding the AI network into a community of practice.
• A separate pilot will map the technology systems councils actually use, to identify common challenges and support longer-term digital transformation.

Councils told the Ministry of Housing, Communities and Local Government exactly where responsible AI adoption hurts: no clear minimum expectations for ethics, guidance that is confusing and fragmented, and too little practical day-to-day support. Those themes, gathered from 60 councils at drop-in sessions in April, have now shaped the Local AI team’s project priorities.

The work is organised around three workstreams — moving from principles to practice, knowledge sharing and peer support, and training and capability building. Headline deliverables include an AI playbook for local government, standard assessment templates, and growing the existing AI network into a community of practice. The team is also aligning with the Local Government Association’s cyber, digital, AI and technology programme for 2026-27.

Nimshi Mahamalimage, AI ethics lead at MHCLG, said the drop-in sessions “were just the starting point”, with co-design workshops to follow for developing and testing the new deliverables.

Why this matters beyond town halls

The complaints councils raised — fragmented guidance, no ethical baseline, insufficient practical support — mirror what UK businesses report about AI adoption generally. The difference is that local government is getting a coordinated central response. Councils buy technology from the same suppliers that sell to small and mid-sized firms, so standard assessment templates and a published playbook could become de facto reference documents well beyond the public sector, much as government security guidance has been borrowed by private-sector buyers.

MHCLG has also launched a pilot to gather detailed information about the technology systems councils run, aiming to spot shared problems and opportunities across the sector rather than leaving each authority to diagnose its own estate.

Looking forward

Co-design workshops will develop and test the deliverables, with progress shared through further blog posts. The practical test is whether the playbook and templates arrive before councils’ AI procurement decisions do — several authorities are already deploying AI in service delivery, and guidance that lands after contracts are signed will have missed its moment.