# Blind political-durability review: entries A-Q ## Scope and method This review uses only the controlled thesis digest, the verified policy fact sheet, the blind-review rubric and anonymised entries A-Q. It gives extra weight to coalition feasibility, sequencing, household agency, rent leakage, democratic legitimacy, reversibility and institutional durability. Confident prose receives no credit where the legal route, administrative base or arithmetic fails. Scores are out of ten on: adoption compatibility (Adopt), observable and administrable bases (Base), UK feasibility (UK), EU and member-state fit (EU), fiscal arithmetic (Fiscal), household distribution and rent defence (House), political acceptability and sequencing (Politics), durability and anti-capture (Durable), bounded regret and triggers (Triggers), and originality and mechanism quality (Original). ## Ranked scorecard | Rank | Entry | Adopt | Base | UK | EU | Fiscal | House | Politics | Durable | Triggers | Original | Total | |---:|:---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:| | 1 | L | 10 | 9 | 9 | 9 | 9 | 9 | 9 | 9 | 9 | 9 | **91** | | 2 | K | 10 | 9 | 9 | 9 | 8 | 9 | 9 | 9 | 9 | 9 | **90** | | 3 | G | 9 | 9 | 8 | 9 | 6 | 9 | 9 | 8 | 9 | 9 | **85** | | 4 | M | 9 | 8 | 8 | 9 | 7 | 8 | 9 | 8 | 6 | 8 | **80** | | 5 | E | 9 | 8 | 7 | 7 | 4 | 8 | 7 | 9 | 7 | 9 | **75** | | 6 | N | 9 | 8 | 9 | 8 | 7 | 6 | 8 | 7 | 6 | 5 | **73** | | 7 | A | 9 | 8 | 7 | 6 | 4 | 7 | 7 | 8 | 6 | 9 | **71** | | 8 | I | 8 | 7 | 7 | 5 | 4 | 8 | 8 | 7 | 7 | 8 | **69** | | 9 | C | 7 | 7 | 7 | 6 | 5 | 8 | 7 | 6 | 7 | 7 | **67** | | 10 | B | 8 | 6 | 6 | 5 | 2 | 6 | 6 | 6 | 6 | 7 | **58** | | 11 | D | 6 | 6 | 5 | 4 | 2 | 5 | 5 | 6 | 4 | 7 | **50** | | 12 | Q | 5 | 5 | 5 | 2 | 2 | 8 | 6 | 5 | 4 | 7 | **49** | | 13 | F | 5 | 5 | 6 | 3 | 2 | 6 | 5 | 6 | 4 | 6 | **48** | | 14 | J | 5 | 4 | 5 | 4 | 1 | 6 | 6 | 4 | 5 | 6 | **46** | | 15 | P | 5 | 5 | 6 | 3 | 3 | 4 | 5 | 6 | 4 | 5 | **46** | | 16 | H | 2 | 1 | 3 | 1 | 1 | 4 | 3 | 2 | 2 | 3 | **22** | | 17 | O | 2 | 1 | 3 | 1 | 0 | 3 | 4 | 2 | 1 | 3 | **20** | The top tier succeeds because it keeps four propositions separate: adverse outcomes may justify preparation; preparation does not itself authorise spending; public investment capacity is not current revenue; and cash should not scale when scarcity rents are absorbing the gain. Lower-ranked entries repeatedly collapse those distinctions. ## Entry-by-entry findings ### A — 71/100 - **Strongest mechanisms:** the Public Upside Rule through distinct support, concession, procurement and tax channels; the two-layer anti-hollowing structure combining territorial claims with upstream participation. - **Weakest mechanisms:** no arithmetic showing that obtainable claims can support material household income; vague automatic fiscal triggers and an overbroad first-parliament agenda. - **Fatal feasibility error:** none if treated as a public-asset acquisition strategy. It is not, however, a funded successor income system as written. - **Unsupported or invented premise:** the Intel transaction, claims of collapsing apprenticeships and vanishing graduate roles, and the claim that open models trail only by months are outside the controlled fact packet and cannot carry the case here. - **Best repair:** enact only a transparently valued public-return rule for new selective support, accumulate realised returns, and add an immediate protected-savings reform before promising citizen dividends. - **Political durability:** “public money earns a public return” has cross-party appeal. The full border-bargaining doctrine invites provider retaliation, Treasury resistance and inter-state undercutting; it should follow, not lead, the coalition sequence. ### B — 58/100 - **Strongest mechanisms:** a cause-neutral labour-share/corporate-margin trigger; combining Universal Credit asset reform with a public capital endowment. - **Weakest mechanisms:** an undefined destination-based economic-value-added tax; an unaffordable fixed dividend presented as a modest anchor. - **Fatal feasibility error:** £1,200-£1,800 per adult would cost roughly £60bn-£100bn annually, far beyond the entry's own £14bn-£16bn surcharge estimate plus slow, uncertain investment returns. - **Unsupported or invented premise:** the £18bn-£21bn payroll-tax cost, £14bn-£16bn surcharge yield and frictionless collection through payment processors are not supported by the supplied facts. - **Best repair:** cap any payment mechanically at independently scored recurring revenue per eligible person and remove guaranteed amounts and endowment yields from the core floor. - **Political durability:** the payroll-relief coalition is plausible, but it would fracture once the missing dividend bill, broad value-added tax incidence and high consumer pass-through became visible. ### C — 67/100 - **Strongest mechanisms:** housing, energy and platform-rent defence as a co-equal pillar; a contingent outcome-based income floor with an explicit no-action state. - **Weakest mechanisms:** a narrow, offshorable compute/energy base; attempting NIC cuts, full-expensing retrenchment, social housing, a capital fund and an income floor in one coalition. - **Fatal feasibility error:** no single fatal error because the entry admits the gap, but the triggered floor has no pre-funded ceiling and cannot safely activate as drafted. - **Unsupported or invented premise:** the £10bn-£15bn NIC cost, £2bn-£5bn compute-levy yield and assertion that reverse-charging imported AI services is readily administrable are unverified. - **Best repair:** begin with scored UC capital/taper reform, support-linked public returns and procurement portability; do not cut NIC until the replacement revenue has been demonstrated. - **Political durability:** the citizen-assembly idea and rent framing help legitimacy. The simultaneous confrontation with investors, data centres, landlords and Treasury makes the initial package too wide. ### D — 50/100 - **Strongest mechanisms:** candidly estimating that the compute levy raises only hundreds of millions; separating levy collection, asset management and disbursement across institutions. - **Weakest mechanisms:** a contribution-policing Participation Income; automatic structural separation and a reserve AI-output tax. - **Fatal feasibility error:** the entry raises Participation Income towards 80% of the National Living Wage despite estimating only £240m-£360m of initial levy revenue, then claims the levy-funded floor remains intact. Automatic structural separation based on concentration alone is also not a lawful substitute for investigation and remedy design. - **Unsupported or invented premise:** the £8bn-£12bn compute-spend base, the proposed EU transition directive route and extension of adjustment funds are not established in the packet. - **Best repair:** delete the 80% income promise, output tax and automatic separation; use any levy only for a capped pilot while existing UC remains the stabiliser. - **Political durability:** conditional “participation” creates a bureaucracy of approved contribution and excludes people precisely when independence matters. The levy-first sequence gives firms a visible cost before households receive a credible benefit. ### E — 75/100 - **Strongest mechanisms:** the correct-channel Public Upside Rule; two-layer anti-hollowing combined with citizen trusteeship, competition and essential-service protection. - **Weakest mechanisms:** no distribution-scale arithmetic; an exceptionally wide agenda without a first-parliament fiscal or coalition cut line. - **Fatal feasibility error:** none if interpreted as a doctrine for acquiring assets and preserving state bargaining power. It cannot show that those assets replace a material fraction of wages. - **Unsupported or invented premise:** several claimed current facts fall outside the controlled packet, including the 70% exposure figure, 40% apprenticeship fall, OBR scenario details, AI Growth Zone savings, specified CMA/DMA actions and external retaliation claims. - **Best repair:** narrow phase one to market-valued claims for actual public support, tested public-contract exit rights and an Economic Agency dashboard; defer the citizen dividend until realised cash exists. - **Political durability:** value for money, operational sovereignty and anti-lock-in can bridge left, centre and national-resilience constituencies. “Own the transition” is persuasive, but compulsory ownership across every lever would unite the Treasury, incumbents and mobile capital against it. ### F — 48/100 - **Strongest mechanisms:** redirecting regressive pension tax expenditure towards broad capital ownership; insisting that public investment buy public upside. - **Weakest mechanisms:** per-AI-assisted-output and AI-revenue contributions reintroduce the measurement trap; the EU account and income architecture exceeds available competence. - **Fatal feasibility error:** a £2,000 account cannot generate a guaranteed £500 annual return without a 25% yield. NWF investment capacity is also used as three-to-five-year bridge funding for current account capitalisation. - **Unsupported or invented premise:** current UK mortgage-interest relief as a major source, the £3bn-£5bn pension-relief yield, EU basic-income authority under Article 153 and ESF+ funding for a permanent experiment are unsupported or wrong. - **Best repair:** convert it into a modest UK young-adult capital pilot funded solely from a scored pension-relief reform, with market returns and no guaranteed distribution. - **Political durability:** ownership rhetoric broadens appeal, but taxing retirement relief to create an unproven state investment account would attract an intense older-saver and financial-sector veto. ### G — 85/100 - **Strongest mechanisms:** a two-key trigger separating structural evidence from fiscal and supply certification; the rent-pass-through pause combined with explicit per-recipient costing formulas. - **Weakest mechanisms:** the illustrative £600 adult/£300 child baseline still implies a very large gross bill without an enacted source; the administrative stack is extensive for a first phase. - **Fatal feasibility error:** none, provided its rule that funding must precede payment is honoured. The statement that the modest floor is plausibly funded is not yet demonstrated. - **Unsupported or invented premise:** no material false current-policy premise identified; its thresholds are clearly labelled as design values. The first-payment funding claim remains unproven rather than factual. - **Best repair:** build the payment rail at zero or pilot scale and begin with protected savings plus earnings insurance until a named recurring Finance Act package has been scored. - **Political durability:** this is one of the strongest coalitions: workers receive insurance, innovators avoid an AI tax, small firms receive interoperability, and fiscal conservatives receive a funding certificate. Universality creates a constituency, but a £30bn-plus gross baseline risks overwhelming that coalition. ### H — 22/100 - **Strongest mechanisms:** recognising that cash needs service and competition protection; proposing audit and multi-stakeholder oversight. - **Weakest mechanisms:** the entire AI-output levy depends on measuring AI utilisation and attributable profit; most institutions, revenues and delivery powers are invented. - **Fatal feasibility error:** the core base directly violates the thesis constraint. The package also allocates £10bn from a levy estimated at £3bn, assumes 15%-20% dividend yields, and creates an EU tax and dividend on an impossible timetable. - **Unsupported or invented premise:** UK and EU AI-use registries, an existing attribution formula, EU payment infrastructure, Help to Buy as a current rent-subsidy vehicle, a UK Skills Funding Agency, AI agencies and risk-sharing exemptions are unsupported or incorrect. - **Best repair:** discard the levy, dividends and institutional timetable; retain only contract-linked portability and a separately funded, tightly defined essential-service pilot. - **Political durability:** “fair share of the boom” is saleable, but the coalition disintegrates when firms must disclose an arbitrary AI share and households discover that the promised benefits are unfunded. ### I — 69/100 - **Strongest mechanisms:** an employer value-added contribution swapped against employer NI; individual Agency Accounts paired with a housing pass-through test. - **Weakest mechanisms:** using NWF capacity as transitional payment support; a legally contested EU tax-coordination directive presented as the year-four route. - **Fatal feasibility error:** the pilot mixes recurring EVAC revenue with NWF “bridging capital”, spending investment capacity as though it were transfer revenue. Article 153 is not a safe route to a binding business-contribution floor. - **Unsupported or invented premise:** the £2.1tn GVA base, £16bn-£21bn EVAC yield and historical tax comparisons are outside the packet; collection through VAT returns is more complex than stated. - **Best repair:** finance a UK-only pilot exclusively from scored EVAC receipts, keep NWF assets on a separate ledger, and treat EU coordination as a non-binding recommendation. - **Political durability:** the NI-for-value-added swap is unusually communicable to SMEs and unions. High-value-added, low-payroll sectors would organise strongly against it, so neutrality and a long transition must be credible before launch. ### J — 46/100 - **Strongest mechanisms:** public equity where public support bears risk; a public-compute outside option for smaller firms and essential services. - **Weakest mechanisms:** a teraflop royalty and foreign-compute border adjustment that are not practically auditable; implausible wealth-fund and levy yields. - **Fatal feasibility error:** a £95.4bn annual grant is supported with invented £18bn-£25bn compute revenue and £12bn annual NWF returns. A £12bn return on a £27.8bn-capacity fund is not remotely conservative, and imported workflow compute cannot be estimated reliably for a border charge. - **Unsupported or invented premise:** every revenue estimate, cloud-margin base, UC saving and public-fund return is outside the controlled facts. - **Best repair:** remove the universal grant and border adjustment; retain a market-valued public-return rule and a small public-compute access programme with ordinary appropriation. - **Political durability:** “Your Dividend from the Machines” is vivid, but an explicit tax on domestic compute creates a powerful investment-flight narrative and the funding failure would destroy trust. ### K — 90/100 - **Strongest mechanisms:** the two-ledger rule separating a tax-funded floor from realised investment dividends; the agency corridor with progressive recovery, funding gates and an essentials absorption brake. - **Weakest mechanisms:** the proposed 0.5%-of-GDP baseline still needs a politically difficult property/capital/consumption package; administration across tax, benefits, housing and investment is substantial. - **Fatal feasibility error:** none. It consistently refuses to spend public-investment capacity or trigger unaffordable cash without a current fiscal score. - **Unsupported or invented premise:** no material invented current fact identified. Proposed percentages and trigger values are appropriately labelled as design ceilings. - **Best repair:** make the first statutory phase payment-rail preparation, protected savings and earnings insurance, with the base credit commencing only on Royal Assent to the named recurring revenue measures. - **Political durability:** this has the strongest complete coalition. It offers firms adoption neutrality, workers visible security, renters a brake, and fiscal conservatives hard ceilings. Announcing the revenue package and credit together is essential to prevent either a “tax first” or “cash first” backlash. ### L — 91/100 - **Strongest mechanisms:** the three-ledger Agency Fund and rent-to-agency waterfall; the two-clock, two-key escalation requiring sustained household deterioration and 110% recurring fiscal coverage. - **Weakest mechanisms:** it deliberately withholds a numeric permanent tax base, which could permit indefinite Treasury delay; the initial package protects agency but is insufficient under a rapid severe break. - **Fatal feasibility error:** none. It is the clearest entry about what current institutions can and cannot fund. - **Unsupported or invented premise:** no material invented present-policy premise identified. It distinguishes supplied facts, proposed thresholds and unknown arithmetic carefully. - **Best repair:** legislate a deadline for Treasury/OBR options and require Parliament to vote on a default broad-base package, preventing “more modelling” from becoming a permanent veto. - **Political durability:** best in the set. It sequences bills, savings, contract exit and visible cost reduction before large cash; it admits Parliament cannot be bound; and it gives local institutions a role without dispersing fiscal accountability. ### M — 80/100 - **Strongest mechanisms:** clean division between national income provision and EU rights/competition; combining a cash floor, protected capital claim and adoption-compatible workplace due process. - **Weakest mechanisms:** a two-quarter disposable-income trigger is too noisy; “modest payment” and broad profit/rent funding remain unpriced. - **Fatal feasibility error:** none. The payment is explicitly conditional on OBR scoring and an identified source. - **Unsupported or invented premise:** no major current-policy invention identified. Thresholds are proposed rather than presented as enacted rules. - **Best repair:** require four warning quarters and eight structural quarters, plus a fiscal certificate and rent-pass-through check before a permanent increase. - **Political durability:** the public case is strong because it adds the right to challenge consequential automated decisions without promising job preservation. High-income taxpayers and rent holders are named honestly, although their organised resistance is understated. ### N — 73/100 - **Strongest mechanisms:** feasible UC asset/earnings reform; disciplined procurement and competition measures using existing competences. - **Weakest mechanisms:** no scalable recurring successor income; adjustment funds and training are treated as the main triggered response even though they cannot replace labour income. - **Fatal feasibility error:** none, but the package openly remains insufficient in the severe thesis scenario. - **Unsupported or invented premise:** using adoption above or below 40% as a decision threshold and front-loading adjustment funds to high-adoption regions are unsupported policy rules, not evidence-backed triggers. - **Best repair:** add a dormant national cash supplement with a recurring funding gate and household outcome trigger, while keeping EU funds confined to eligible adjustment uses. - **Political durability:** its modesty is its strength: fiscal conservatives, business and unions can all accept the first phase. Its weakness is motivational; diffuse lower prices and better procurement may not create a constituency capable of defending the architecture. ### O — 20/100 - **Strongest mechanisms:** a universal individual payment is easy to explain; a gross-revenue base is less vulnerable to profit shifting than local subsidiary profit. - **Weakest mechanisms:** the AI-output share is unmeasurable; the rent cap, EU revenue pool and transition eligibility all depend on fictional definitions and powers. - **Fatal feasibility error:** the arithmetic is wrong by three orders of magnitude: 67 million multiplied by £1,200 is £80.4bn, not £80.4m; 450 million multiplied by €1,200 is €540bn, not €540m. It also assumes an EU tax and dividend through qualified majority. - **Unsupported or invented premise:** a 12% average AI-revenue share, £4.5bn levy yield, existing EU digital revenue pool, EU-wide dividend authority and 67 million UK adults are invented or false. - **Best repair:** no incremental repair is sufficient. Rebuild around broad national tax bases, correct population arithmetic and member-state transfers without any AI-share test. - **Political durability:** the £1,200 headline would initially be popular, but the inevitable fiscal exposure would make the entire settlement synonymous with deception. ### P — 46/100 - **Strongest mechanisms:** contract-linked portability and transparent audit; candid recognition that the compute contribution yields negligible revenue. - **Weakest mechanisms:** the service basket omits the central housing and energy rent problem; a job guarantee and training accounts do not create independent post-wage agency. - **Fatal feasibility error:** the EU tax and compute contribution routes are not legally available as described, and ESF+ cannot finance a standing statutory job guarantee. Proposed PLTA and job-guarantee costs exceed the stated revenue. - **Unsupported or invented premise:** UK/EU digital-revenue bases, service programme costs, CPU-hour totals and an existing EU DST directive are unsupported. - **Best repair:** discard the compute contribution and job guarantee; retain procurement clauses and fund a narrower broadband/transition pilot through an ordinary scored national tax measure. - **Political durability:** the green-labour coalition is identifiable but narrow. Fiscal moderates and innovators receive little protection from an expanding digital tax, compute charge and employment guarantee. ### Q — 49/100 - **Strongest mechanisms:** unusually concrete housing, land and platform-rent measures; honest disclosure of very large UK and EU dividend gaps. - **Weakest mechanisms:** the timetable spends dividends before returns exist; the EU legal architecture relies on contested or unavailable tax and social-policy bases. - **Fatal feasibility error:** the transition borrows against uncertain future levies and schedules dividends from speculative equity returns. Article 114 cannot safely create the proposed compute tax, Article 153 cannot straightforwardly create the dividend and basic-services settlement, and enhanced cooperation does not solve the own-resource problem. - **Unsupported or invented premise:** data-centre counts, platform revenues, capital-tax yields, UC savings, four-to-five-percent public returns and first-dividend amounts are unverified. - **Best repair:** retain only UK capital-income/property reform, public-support equity and a housing pass-through pilot; pay no compute dividend until audited cash returns exist. - **Political durability:** the renter/platform coalition is real, but the plan simultaneously confronts landlords, financial capital, data centres, platforms and low-tax states while promising benefits the revenue cannot cover. ## Convergence map | Recurring mechanism | Entries | Political and feasibility reading | |---|---|---| | Public return when public capital or selective support bears risk | A, B, C, D, E, F, G, H, I, J, K, L, M, N, Q | The strongest genuine convergence. It is durable only when valuation is commercial, procurement remains separate and realised returns are not booked in advance. | | Individual cash or capital agency account | A-Q | Universal convergence in aspiration, not in funding. Entries K and L provide the safest accounting; O and H demonstrate the fiscal failure mode. | | Competition, portability and public-contract exit rights | A-Q | Very strong low-regret convergence. Contract nexus and case-specific remedies are essential. This protects income and state capacity but does not itself create household income. | | Housing, energy or platform rent defence | A-E, G-M, N, P, Q | Strong convergence, uneven quality. K, L and G are strongest because rent evidence pauses cash scaling rather than merely adding a slogan. | | Outcome dashboard and dormant triggers | A-Q | Universal convergence. Best versions use several household/labour measures plus a fiscal certificate; weak versions trigger on AI adoption, two noisy quarters or self-reported AI output. | | UC capital-cliff/protected-savings reform | B, C, D, G, I, J, K, L, M, N, Q | Strong practical convergence and the quickest household-agency win. Cause-neutral design is more legitimate than occupation or AI-displacement tests. | | Shift financing away from payroll alone | A-Q | Broad convergence, but no consensus base. Property, land, capital income and destination consumption are more durable than AI, token, output or domestic-compute bases. | | AI, compute or digital-sector levy | C, D, F, H, I, J, O, P, Q | Recurrent but weak convergence. Most versions have import, avoidance, siting or attribution failures; several directly violate the thesis constraint. | | Cause-neutral earnings insurance | G, K, with partial analogues in L and M | Less frequent but politically powerful: immediate, bounded and compatible with adoption. It should bridge transition rather than claim to replace the long-run circuit. | | Universal basic services or essential-cost provision | A, C, E, G, H, J, K, L, M, N, P, Q | Important anti-leakage convergence. Housing supply and local infrastructure are more credible than broad rent caps or unfunded service entitlements. | Convergence is not proof. Fifteen entries can repeat public equity while still failing to show distributable yield; all entries can repeat a citizen account while only a few identify the fiscal boundary. ## Three strongest mechanisms across the set ### 1. The outcome-and-funding escalator with an absorption brake The best design, clearest in G, K and L, requires both sustained deterioration in several economic-agency measures and independent certification of recurring funding, delivery readiness and rent effects. Statistical breach authorises consideration; it does not create unlimited spending. Cash increments pause when housing, energy or platform costs absorb the gain. ### 2. Public return for public risk, with correct channels and ledgers When government contributes capital, guarantees, concessional finance or a scarce selective advantage, it should seek a proportionate, valued equity, convertible, royalty, repayment or revenue claim. Procurement should separately secure contract-related portability, audit and exit rights. Capital, cost-reduction investment and recurring cash must sit on separate ledgers; only realised net returns can fund a dividend. ### 3. Immediate cause-neutral household agency Replace the Universal Credit capital cliff with a protected reserve and taper, and pilot capped declining earnings insurance triggered by verified earnings loss rather than an AI-causation test. This gives workers, carers, freelancers and small-business founders a real exit option now, is reversible and creates a coalition before any large permanent transfer is affordable. ## Three most common failure modes 1. **Fiscal category errors and impossible arithmetic.** Entries spend investment capacity as revenue, treat mobilisation totals as cash, assume double-digit or 25% returns, or promise universal payments one to three orders of magnitude above their stated sources. 2. **Technology-attribution and narrow-base failure.** AI-output shares, “AI-adjusted profit”, autonomous outputs, teraflop border estimates and AI-assisted service revenues are gameable. Domestic compute and data-centre energy are observable locally but mobile, narrow and prone to cost pass-through or relocation. 3. **Legal and constitutional fantasy.** Repeated errors include EU taxes or dividends by qualified majority, Article 153 as a general income or tax power, ESF+ as a permanent entitlement, unrelated procurement equity, automatic structural separation, and UK supermajorities portrayed as binding future Parliaments. Two secondary failures repeatedly damage political durability: triggers fire on noisy or adoption-based indicators, and “rent defence” is asserted without a deliverable housing-supply or essential-cost sequence. ## Recommended integrated package: the Agency Security Compact The politically durable package should be materially narrower than the union of proposals. It should not begin with a universal dividend, compute tax, EU own resource, job guarantee or new central mega-fund. ### Phase 1: household security and state capability, months 0-18 1. **Protected savings and cause-neutral transition insurance.** Replace the UK Universal Credit £16,000 cliff with a protected transition reserve and tapered capital treatment, subject to OBR scoring. Pilot capped declining earnings insurance for up to 18 months using verified tax and payroll losses. Do not require proof of AI causation. Preserve disability, child and housing support separately. 2. **Economic Agency Accounts.** Publish labour compensation share, paid hours per working-age adult, median market income relative to productivity, wage-financed consumption, entry-level hiring, disposable resources after housing and energy, and tax-base composition. Freeze definitions, show revisions and avoid a single AI-displacement statistic. 3. **Contract exit and competition.** Put testable portability, interoperability, evaluation, audit, security, step-in and supplier-exit clauses in relevant public contracts. Use CMA and DMA powers case by case against lock-in and data barriers. Do not demand unrelated parent-company equity or domestic preference. 4. **Public-return rule.** Apply only to new transactions where government actually bears commercial risk or grants a scarce selective advantage. Publish the benefit supplied, valuation, instrument, expected return, subsidy/State aid analysis and exit rationale. Use equity, convertibles, royalties or repayable finance as the transaction warrants. ### Phase 2: three ledgers and a funded payment rail, months 12-30 Create an independently audited Agency Fund with: - a **capital ledger** for equity, convertibles and one-off receipts, retaining loss reserves; - a **cost-reduction ledger** for housing supply, local infrastructure, energy efficiency and digital access with measured household outcomes; and - a **recurring agency ledger** containing only enacted, scored recurring revenue and realised net cash returns. Build an individual payment rail and progressive tax-recovery mechanism, but initially pay nothing beyond a tightly funded pilot. The first national Agency Credit begins only when Parliament has enacted named recurring revenues and the OBR certifies coverage. Candidate bases should be broad and auditable: high-value immovable property or land, better alignment of labour and capital-income treatment, realised profits and distributions, and, if necessary, a destination-based consumption contribution recycled progressively. No AI-specific levy is required. ### Phase 3: two-key scaling, years 3-5 A warning review begins when at least three of five pre-set indicators remain materially below trend for six quarters, including either labour share or median market income relative to productivity. A permanent increment requires: - persistence for eight quarters; - independent confirmation that the change is not a data break or ordinary short recession; - recurring five-year funding coverage of at least 110% under stress; - benefit-delivery readiness; and - a housing, energy and platform-rent pass-through assessment. The first net credit should be capped at the lower of enacted recurring revenue and a deliberately small fiscal envelope. Later steps require a fresh parliamentary vote and current distributional score. Temporary supplements may draw only on a pre-funded reserve and expire automatically. If fewer than two indicators remain breached for eight quarters, temporary tiers phase down prospectively; payments already made are never reclaimed. ### EU and member-state division - **EU institutions:** common measurement, AI-rights implementation, competition, portability, procurement guidance, State aid scrutiny and transparent public-investment terms. - **Member states:** protected savings, earnings insurance, recurring revenue, cash payments, housing and energy policy. - **No near-term EU tax or dividend:** prepare options only if unanimity, national ratification and a funded budget route actually exist. ### Coalition and democratic legitimacy The coalition should be built around immediate, reciprocal benefits: - Workers, carers and freelancers receive savings protection, transition income and due process. - Innovators and SMEs receive no robot tax, no job quota and better switching/public-compute access. - Fiscal conservatives receive hard expenditure ceilings, independent scoring, sunsets and no speculative investment income. - Renters and bill payers receive a pass-through brake and visible cost-reduction investment. - Taxpayers receive public upside when public capital is at risk. The principal losers are owners of high-value immovable property, recipients of substantial lightly taxed capital income, gatekeepers profiting from lock-in and recipients of gratuitous public support. Legitimate liquidity problems should receive deferral with interest, not permanent exemption. Rules should phase in, avoid retrospective confiscation and simplify compliance for small organisations. The public case should fit one sentence: > Use the technology; protect people's savings; make public risk earn a public return; expand cash only when the evidence, funding and essential-cost capacity are real. This package preserves democratic choice. It does not pretend a future Parliament can be bound. Durability comes from visible individual benefits, separate audited ledgers, explicit votes, distributed institutional responsibility and a coalition that has something concrete to lose if the architecture is raided. ## Final judgment Entry L is the strongest political-durability design, narrowly ahead of K. G supplies the best worked trigger and rent-pass-through mechanics, while A and E supply the strongest public-upside and anti-hollowing doctrine. The integrated recommendation takes those mechanisms but rejects their unfunded or overbroad extensions. The council's genuine consensus is not “tax AI and pay a dividend.” It is narrower and more defensible: stop punishing precautionary savings, measure household agency rather than AI labels, make public risk earn a public return, keep critical markets switchable, prevent scarcity from swallowing support, and require both evidence and recurring finance before permanent cash expands.