An open dialogue with public institutions
42 of the gaps on this map need no new law, just a willing institution at the table. This page is our standing offer: for each one, what we’re asking, what we bring, what you gain, and a first step small enough to agree in one meeting. If you work inside one of these institutions, this is the agenda.
These are written for both sides of the table: the gain named for your institution is meant honestly, and the first steps are deliberately pilot-sized, time-boxed and reversible. Every brief links to a full dossier with sources.
Local government (10)
No grant-based retraining instrument for mid-career workers displaced by AI
The ask: Pilot grant-funded retraining accounts for workers in high-AI-exposure occupations, paid from the adult skills budget you already control.
First step: A twelve-month pilot of 200 grant accounts in one authority, capped at existing bootcamp unit costs, jointly evaluated, with a stop-go review at six months.
Planning reform without planners: no workforce pipeline at the scale required
The ask: Fund shared specialist planners and apprenticeships with two or three neighbouring councils, using the full-cost fees the Planning and Infrastructure Act already lets you set.
First step: A six-month trial: three neighbouring districts share one ecologist and one heritage officer, costs recovered through planning fees, workload and decision times reviewed together after two quarters.
No standing mass-transit delivery unit or rolling tram programme for second cities
The ask: Sponsor, with fellow mayoral authorities, a shared light-rail delivery unit: one standard specification, one procurement framework, one engineering team serving every city's scheme.
First step: A six-month joint feasibility study between two authorities, funded from existing programme budgets, testing whether their next tram schemes could share a specification and procurement framework, with no commitment beyond the study.
England has no co-operative development agency
The ask: Franchise a pilot regional co-operative development agency within your combined authority, offering formation advice and finance brokerage to new co-ops and community businesses.
First step: A six-month desk inside your existing business-support service: one co-operative development adviser, funded by us, jointly evaluated on enquiries handled and co-ops formed, creating no new institution and ending cleanly if it fails.
No regulatory pathway or public infrastructure for mutual credit and B2B obligation-clearing
The ask: Anchor a city-region invoice-clearing pilot, letting your local suppliers offset debts against each other and easing their cash flow without bank credit.
First step: Six months as named observer and endorser of a clearing club for fifty local suppliers, with a published joint evaluation; whether the council itself trades is a later, separate decision.
Community mesh communications absent from UK civil resilience doctrine
The ask: Adopt a neighbourhood radio-mesh pilot in your emergency plan: a practised text fallback for when power and phone masts fail.
First step: A 90-day trial in one parish: twenty nodes across halls, shops and volunteers' homes, one exercise simulating a mast outage, and a one-page joint evaluation for the resilience forum.
No place-based transition plans for AI-exposed local economies (contact-centre towns)
The ask: Work with us on the country's first AI transition compact: exposure map first, then pooled retraining money and early-notification arrangements with large local employers.
First step: A 90-day exposure-mapping exercise for the authority's area, using public data, presented to officers before anything is published, proceeding to a compact only if the map shows a case.
Local Public Accounts Committees: promised exploration, still no place-based spending scrutiny
The ask: Host a two-area pilot of a local public accounts committee, funded by philanthropy, to follow the pound across everything spent in your area.
First step: A six-month shadow committee tracing one programme's money across council, mayoral and NHS budgets, with an independently chaired public session and a single published report, at our cost, evaluated jointly.
No national civil-protection volunteer reserve
The ask: Bring a trained, insured pilot volunteer corps into your local resilience forum's plans and exercises.
First step: Invite the pilot corps into one planned exercise this season under a simple agreement covering vetting and insurance; review together afterwards before deciding on any standing role.
No verifiable spending and procurement rails for local government
The ask: Publish one service's payments and procurement as they happen, in a tamper-evident open format, and evaluate the results with us.
First step: A 90-day pilot publishing one service's spend in the open format alongside your existing files, jointly evaluated; if it adds work without value, switch it off.
NHS & care (2)
No institutional home for open code in the NHS, and an active retreat from it
The ask: Work with one volunteer trust or care board to trial open-code and open-API procurement conditions, as the first step to reinstating a national open source policy.
First step: One trust or care board applies open-code conditions to a single non-critical procurement over six months, with the security case documented and reviewed jointly before anyone talks about national policy.
No credible sovereign exit option for the NHS Federated Data Platform before the February 2027 break clause
The ask: Commission an independent assessment of what leaving the Palantir platform would take, before the February 2027 break clause forecloses the choice.
First step: A 90-day exit-feasibility assessment, independently published, alongside one willing trust scoping an open-standards pilot for a single operational workflow; neither commits anyone to leaving.
Schools & education (3)
England has no national school digital platform for safe central provision of AI tools (no Hwb equivalent)
The ask: Form a founding coalition of school trusts to build a federated framework (single sign-on and jointly negotiated licences for vetted AI tools), open to any school that joins.
First step: A two-term pilot: five trusts jointly negotiate one AI tool licence and trial shared sign-on across ten schools, comparing cost and safeguarding overhead with buying alone.
No dedicated national R&D programme on AI pedagogy and child-AI interaction
The ask: Join a volunteer network of schools running independent trials of the AI tools you are being asked to adopt, before they reach every classroom.
First step: A two-term trial in ten to twenty volunteer schools of one government-endorsed tutoring tool, findings published whatever they show, any school free to withdraw at any point.
Home-educated candidates have no guaranteed route to public examinations
The ask: Build a guaranteed examination route for home-educated candidates with us: exam boards accrediting an access network, centres offering places at cost.
First step: One exam season in a single county: three centres offer a fixed block of private-candidate places at audited cost, the charity handles bookings, and results and costs are published jointly.
Regulators (1)
Unions & membership bodies (3)
No funded national programme for AI-era assessment redesign in higher education
The ask: Turn the one-year national assessment pilot into a standing shared programme: pooled assessment banks, shared oral-exam capacity and a common standard for verifying degrees, built with volunteer universities.
First step: A one-semester pilot: five member universities pool redesigned assessments in two high-risk disciplines through the existing pilot's infrastructure, with uptake and integrity results shared across the membership.
No funded national capability programme for council finance professionals
The ask: Run a 'Teach First for treasurers' pilot with us: one bursaried graduate cohort placed in volunteer councils, qualifying through your route.
First step: One cohort: ten graduates, five volunteer host councils, seed-funded for two years to part-qualification, with a joint review after year one before anyone commits to a second.
No civic governance laboratories: verifiable voting and decision-records untested in low-stakes democracy
The ask: Run a verifiable ballot alongside one of your existing votes: the incumbent result stands; the trial simply builds the evidence.
First step: One consultative ballot or annual-meeting motion run in parallel with your normal process over three months; members compare both, and the normal result is the only one that counts.
Funders & foundations (4)
No acquisition capital behind the new Community Right to Buy
The ask: Build with us a permanent acquisition fund (grants, long-term loans and match funding for community shares) that can back a bid within the statutory moratorium window.
First step: A twelve-month pilot: one community foundation underwrites up to ten moratorium-window bids in two regions, capped at £2 million, independently evaluated before any wider commitment.
No transition mechanism turning closing faith buildings into community assets
The ask: Agree standard legal templates and a shared feasibility fund so closing places of worship can pass into community hands instead of dereliction or conversion to flats.
First step: A twelve-month pilot in one diocese: three closing buildings taken through a templated transfer with small feasibility grants, jointly reviewed before anything is standardised nationally.
No transition instrument for AI-displaced workers: the UK has no TAA equivalent
The ask: Act as funder-of-record for a pilot transition fund pairing time-limited wage insurance with a retraining entitlement, starting in occupations already contracting.
First step: Co-fund a six-month design phase: one occupation, one town, costed per participant, ending in a go or no-go decision that binds nobody to the full pilot.
No payout floor or distribution transparency for endowed foundations
The ask: Convene the largest grantmakers into a voluntary pledge to publish their payout ratios annually, in one shared format, before anyone designs a statutory floor.
First step: A twelve-month pilot: twenty volunteer foundations publish one year's payout ratio in the shared format, compiled and released jointly, with any foundation free to withdraw.
Government bodies (17)
No evidence institution for social connection, loneliness and civic health
The ask: Co-fund a What Works Centre for social connection and civic life, endowed for ten years and formally owning the Know Your Neighbourhood evidence base.
First step: A six-month scoping study, jointly funded at under £100,000, testing demand and a hosting model: the same first step every previous What Works Centre took.
No official statistics on social infrastructure and associational life
The ask: Work with us to develop official annual statistics on community spaces and participation at neighbourhood level, starting from an open prototype rather than a blank page.
First step: A 90-day joint feasibility review of the prototype against official-statistics standards, published as a methodology note, with no commitment to badge anything beyond that.
No participation pipeline: volunteering infrastructure and a successor to NCS
The ask: Pilot a volunteering passport with us (one enhanced check, portable between organisations) and co-design a voluntary civic year delivered through existing youth organisations.
First step: A six-month pilot in one county: volunteers from three youth organisations share one enhanced check under existing update-service rules, with drop-out and safeguarding outcomes measured jointly.
No national assurance service giving public bodies confidence in open source products
The ask: Turn the Sourcing the Stack criteria into a standing assurance service by piloting kite-marked evaluations of open source products councils already want to buy.
First step: A 120-day pilot: evaluate three open source products against the seven draft criteria with volunteer reviewers, publish the results openly, and gather council feedback before deciding anything about a permanent service.
No census of the critical open source dependencies underpinning UK government and CNI
The ask: Commission a recurring census of the open source software that government services and critical national infrastructure actually run on, ranked by criticality and maintainer health.
First step: A 90-day pilot aggregating dependency data from five volunteer digital services, producing a criticality ranking for internal review only, and a joint decision afterwards on whether a full census is worth commissioning.
No workforce pipeline to replace 4,500 lost qualified youth workers
The ask: Fund a national youth work workforce programme (bursaries, degree apprenticeships and a graduate fast-track) as the missing delivery plank of the National Youth Strategy.
First step: A twelve-month co-funded pilot: fifty bursaried trainees placed with existing youth work employers in two regions with waiting lists, retention and cost per qualification evaluated jointly against teaching bursary benchmarks.
No official statistical series or open dataset on youth participation and provision
The ask: Commission an annual official measure of youth civic participation, including eleven-to-fifteen-year-olds, and publish an open national dataset of youth provision alongside it.
First step: A 90-day joint exercise merging our shadow dataset with departmental returns for one region, published in an open format, with coverage assessed together before any decision on a national series.
No independent pre-deployment testing infrastructure for police surveillance algorithms
The ask: Host a pilot independent testing facility for police facial-recognition and algorithmic tools, publishing accuracy and demographic error rates before any force deploys them.
First step: A six-month scoping study: agree a test protocol modelled on the American programme, evaluate one vendor's algorithm on a voluntary basis, and publish the method and results jointly.
No UK public-sector or civic deployment of sovereign messaging (a UK 'Tchap')
The ask: Pilot a Matrix-based government messenger for one team, as France, Germany and more than twenty-five other countries already run.
First step: A 90-day trial server for one non-sensitive team of fifty, run alongside existing tools, evaluated against cost, usability and records compliance, and switched off without ceremony if it fails.
No community access tier for public compute and open-weight AI
The ask: Add a community tier to public AI compute: hosted open-weight models and hands-on support for accredited charities, advice services and local newsrooms.
First step: One pilot cohort: ten accredited organisations, a fixed slice of existing compute for six months, one shared support contractor, and a jointly published report on costs and outcomes.
No public funder for PETs deployment or privacy infrastructure maintenance
The ask: Run one privacy-enhancing-technologies challenge with us: a real data-sharing problem from your portfolio, solved with techniques that never expose the underlying data.
First step: A 120-day pilot on one existing, stalled data-sharing use case: define the privacy constraint, run the challenge, publish the evaluation jointly. No procurement commitment beyond the pilot.
No real-time observatory of AI-driven labour displacement
The ask: Open an accredited-researcher route to link payroll records, job adverts and firm-level AI adoption, and build quarterly displacement indicators together.
First step: One Secure Research Service project, six months: link payroll data and job adverts for three contracting occupations, findings reviewed jointly before publication. Existing access rules; nothing new signed.
No core-funding backstop for evidence infrastructure (What Works Centres)
The ask: Design with us a pooled ten-year core-funding facility for cross-departmental evidence bodies, capitalised jointly by government and foundations.
First step: A jointly written options paper (three funding models, costed, foundation interest tested in principle), finished within ninety days and committing no money from anyone.
No at-scale commercial and procurement capability service for local government
The ask: Pilot a deployable commercial team with us (no-fee support for councils facing reorganisation contract work) as the working case for a national service.
First step: One region, six months: a grant-funded deal team supports three councils through reorganisation contract reviews at no charge, with take-up and results reported jointly.
No funded, standing protocol to mobilise civil society in emergencies
The ask: Adopt a standing protocol giving charities and community groups pre-agreed emergency roles, with a surge fund that philanthropy pools first.
First step: A 90-day trial in one resilience forum area: the draft protocol tested in a single exercise, the fund making one simulated rapid payment, findings published jointly.
No auditable records infrastructure for guardianship and deprivation-of-liberty decisions
The ask: Pilot tamper-evident decision logs on one region's deputyship caseload, alongside the redesign of the liberty safeguards.
First step: A six-month pilot in one region: new deputyship decisions logged in the tamper-evident format alongside existing records, access limited to your staff and consenting families, jointly evaluated before any wider use.
No human-rights due-diligence standard in public technology procurement
The ask: Adopt a published human-rights due-diligence standard for technology procurement in one authority's tenders: a model questionnaire and risk framework used as conditions of participation under the existing Act.
First step: One willing contracting authority pilots the questionnaire on a single technology tender and publishes what it learned.
Electoral bodies (2)
No at-scale 'first vote' programme for 1.6 million newly enfranchised 16-17-year-olds
The ask: Scale your seven democratic education pilots into a guaranteed first-vote induction, with a mock election, in every school and college before the next general election.
First step: A one-term pilot in one local authority: every school and college offered a provider-delivered induction and mock election, with voter registration measured jointly against neighbouring areas.
No decentralised attestation network for critical public facts
The ask: Co-sign one election's declared results cryptographically: an added layer of proof on figures you already publish.
First step: At one council area's next scheduled elections, co-sign declared results in parallel with normal publication; the legal declaration is untouched, and we jointly review who used the verification and how.
How to start
Reach us through the repository. Officials: name the gap number and we’ll bring the dossier and the people. Community groups: tell us which institution you already have a relationship with; warm introductions move faster than cold letters. Nothing here requires anyone to commit beyond the first step.