Beyond Universal Basic Income

  1. Executive Summary

Artificial intelligence is now the fastest-growing single cause of corporate layoffs. For the first time in the history of Challenger, Gray & Christmas’s tracking of U.S. job cuts, AI was the most-cited reason for layoffs in a single month in early 2026^[1]. Goldman Sachs estimates that 6–7% of the U.S. workforce could be displaced during a roughly ten-year AI transition, with around 300 million jobs exposed globally^[2]. As the prospect of structural technological unemployment moves from theory to policy debate, governments are openly discussing Universal Basic Income as a backstop^[3].

This paper argues that UBI, while well-intentioned, is the wrong instrument for AI-driven redundancy. It is universal in targeting the problem, taxpayer-funded when the gains are corporate, and unconditional when the policy aim is re-employment. We propose an alternative we call the Continuity Dividend: a legal obligation for any corporation that eliminates a role through AI to pay the displaced employee 75% of their former salary, indexed and continuing until they find comparable alternate work — or, in defined circumstances, for life.

Our central finding is that the Continuity Dividend is not merely affordable — it is self-funding from the AI savings that cause the redundancy. Removing a fully loaded role typically saves an employer around 142% of base salary[4]; paying 75% of that back to the worker still leaves the firm structurally ahead, even after accounting for the cost of the AI systems that replace the role. At a national scale, the scheme costs roughly 0.34% of GDP — about one-eighth the cost of a poverty-line universal UBI.

  2. Universal Basic Income: Definition and Origins

2.1 What UBI Is

Universal Basic Income is a policy under which the state pays every citizen a regular, unconditional cash sum, regardless of employment status, wealth, or willingness to work^[5]. Its four defining features are that it is universal (paid to all), unconditional (no means test or work requirement), individual (paid to people, not households), and regular (a predictable, recurring payment rather than a one-off grant).

2.2 Historical Roots

The idea is old. It traces to the sixteenth century — Thomas More’s Utopia (1516) and the proposals of Juan Luis Vives for a guaranteed subsistence minimum^[6]. It re-emerged through thinkers from Thomas Paine to Milton Friedman (via the negative income tax), and was tested in the 1970s Canadian “Mincome” experiment, where most recipients kept working, and hospitalisations and mental-health admissions fell^[7].

2.3 Why AI Revived the Concept

The modern revival is driven explicitly by automation anxiety. Andrew Yang’s 2020 U.S. presidential campaign placed a $1,000-a-month “Freedom Dividend” at its centre, arguing that AI and automation would displace so many workers that UBI was necessary to protect the economy^[8]. The logic is straightforward: if AI permanently removes the need for large categories of human labour, then income must be decoupled from work. Pilot programmes in Finland, Kenya, the United States, and Catalonia have since tested versions of the idea, generally improving well-being and economic security while raising unresolved questions about cost, inflation, and work incentives^[9].

  3. The Case For and Against UBI

The evidence base is genuinely mixed. The table below summarises the strongest arguments on each side, drawn from pilot data and economic analysis.

Argument FOR	Why it matters	Argument AGAINST	Why it matters
Reduces poverty & inequality	Pilots show clear gains in income security and physical/mental health; a poverty-line UBI could cut UK poverty rates by ~72%.	Universality is inefficient	Paying everyone — including the wealthy — dilutes funds; targeting the poor directly can do more per dollar spent.
Provides a stable safety net	A predictable floor lets people absorb shocks, retrain, or start businesses without destitution.	High headline cost	A poverty-line UBI for all adults is estimated at ~2.67% of U.S. GDP (~$784B/year), straining public finances.
Supports entrepreneurship & care	Recipients take more risks, pursue education, and value unpaid care and family work.	Possible work disincentive	Critics argue unconditional cash can reduce labour-force participation, though pilot evidence is weak and contested.
Administratively simple	A single universal payment can replace a patchwork of means-tested programmes and their overhead.	Inflation risk	Injecting broad cash demand may raise prices, though the financing method matters and evidence is inconclusive.
Decouples survival from a shrinking job market	Directly addresses the AI scenario in which there may simply not be enough paid work for everyone.	Funded by taxpayers, not beneficiaries of AI	The public pays for a problem created by, and enriching, the firms automating the work — a moral-hazard mismatch.
Administratively simple	A single universal payment can replace a patchwork of means-tested programmes and their overhead.	Inflation risk	Injecting broad cash demand may raise prices, though the financing method matters and evidence is inconclusive.
Decouples survival from a shrinking job market	Directly addresses the AI scenario in which there may simply not be enough paid work for everyone.	Funded by taxpayers, not beneficiaries of AI	The public pays for a problem created by, and enriching, the firms automating the work — a moral-hazard mismatch.

The decisive objection, for the purpose of AI-driven redundancy, is the last row. UBI asks the general taxpayer to fund the consequences of a transition whose profits accrue overwhelmingly to a small set of corporations. McKinsey estimates generative AI could add $2.6–$4.4 trillion annually to global corporate profits^[10]. A policy that socialises the losses while privatising those gains is politically fragile and ethically incoherent. This is the gap the Continuity Dividend is designed to close.

  4. The Alternative: The Continuity Dividend

4.1 The Core Proposition

We propose a simple, enforceable rule. When a corporation eliminates a role primarily because that role’s work has been automated by AI, the corporation must pay the displaced employee 75% of their prior base salary, paid monthly, continuing until the employee secures comparable alternate employment. Where no comparable work is ever found — for example, for older workers near the end of their careers — the payment continues for life, functioning as an AI-funded early pension.

4.2 Why It Is Called the "Continuity Dividend"

The name is deliberate and does two jobs. Continuity signals that the worker’s income, dignity, and consumer spending power continue uninterrupted through the transition — the policy preserves economic continuity rather than triggering a cliff-edge. Dividend reframes the payment not as a penalty or a tax, but as the worker’s share in the productivity windfall that their automation created. The employee is, in effect, a stakeholder receiving a dividend on the value their former role generated. This framing is central to corporate buy-in, discussed in Section 7.

4.3 Design Principles

• Polluter-pays, not taxpayer-pays. The firm that captures the AI saving funds the human cost it creates.
• Targeted, not universal. Only genuinely AI-displaced workers receive payments, making the scheme far cheaper than UBI.
• Pro-work, not anti-work. Payment tapers to zero on finding comparable work, with earnings top-ups (Section 6) so re-employment always pays.
• Dignity-preserving. 75% maintains a near-full standard of living, protecting both the worker and aggregate consumer demand.

  5. Proving Feasibility: The Numbers

The central test of any redundancy model is whether the entity required to pay can actually afford it. We prove feasibility at two levels: the unit economics of a single displaced worker, and the macroeconomic cost at national scale. All figures are illustrative but built on documented benchmarks.

5.1 Unit Economics: One Displaced Worker

Consider a white-collar role with a base salary of $80,000. The employer’s fully-loaded cost — wages plus payroll taxes, benefits, and overhead — is approximately 1.42× base salary, or $113,600^[11]. When AI eliminates the role, the firm stops paying this entire amount but incurs a new cost for the AI tooling that performs the work. The table below shows the firm remains substantially ahead even after paying the 75% Continuity Dividend.

Line item	Amount (per worker/yr)	Note
Fully-loaded cost of the human role (1.42 × $80,000)	$113,600	Cost removed by automation
Less: cost of AI tooling to perform the work (~10% of salary)	– $8,000	New recurring software/compute cost
Net gross saving to the firm	$105,600	Before any obligation to the worker
Less: Continuity Dividend (75% of $80,000)	– $60,000	Paid to the displaced worker
Net annual gain retained by the firm	$45,600	Firm is still ~40% ahead vs. keeping the role

Result: The firm keeps roughly $45,600 per displaced worker every year — about 40% of the original labour cost — while the worker retains 75% of their income. The model is self-funding by construction. In fact, the AI tooling could cost up to 67% of salary before the firm reaches break-even, leaving a wide margin of safety.

5.2 The Re-employment Taper

Because payments stop when the worker finds comparable work, the firm’s actual liability is usually short-lived. Wage-insurance research shows most displaced workers are re-employed within one to two years, and that re-employment support measurably shortens the jobless spell^[12]. Over a typical one-to-two-year claim, the firm’s cumulative net gain ($45,600–$91,200 per worker) comfortably exceeds its cumulative payout. The lifetime-payment provision applies only to the minority who never find comparable work — a tail risk that pooling (Section 6.3) is designed to absorb.

5.3 Macroeconomic Affordability

Scaling to a national economy confirms the model is modest in aggregate. Using Goldman Sachs’s estimate that ~6.5% of the U.S. workforce (about 10.9 million workers) is displaced over a ten-year transition^[13], and assuming an average claim duration of 1.5 years, the steady-state number of concurrent claimants is roughly 1.64 million.

Macro metric	Value
Concurrent claimants (steady state)	~1.64 million
Average Continuity Dividend (75% of $80,000)	$60,000 / year
Aggregate annual payout	~$98 billion
As share of conservative U.S. AI profit uplift (~$1T)	~9.8%
As share of U.S. GDP	~0.34%
Cost of poverty-line universal UBI (for comparison)	~2.67% of GDP
Continuity Dividend vs. universal UBI	~8× cheaper

Conclusion: A scheme costing ~0.34% of GDP, fully funded by the corporations capturing AI gains, is demonstrably workable. It is roughly eight times cheaper than universal UBI precisely because it pays only those actually displaced, and it places the cost where the benefit lands.

  6. Government Mechanisms for Success

A corporate obligation only works if the government builds the scaffolding to define, enforce, and guarantee it. We recommend six mechanisms.

6.1 A Legal Definition of "AI-Caused Redundancy"

Legislation must define when a redundancy is AI-attributable, to prevent both evasion and over-claiming. A practical test: a role is AI-displaced if, within a defined window (e.g. 12 months) of termination, its core tasks are performed by an AI system the employer has deployed. The burden of proof sits with the employer, who holds the deployment records.

6.2 A National Continuity Registry and Levy Authority

A government body — the Continuity Authority — registers every AI-attributable redundancy, verifies payments, and administers disputes. Employers report displacement events as much as they report payroll, creating an auditable national dataset on AI labour impact.

6.3 A Guarantee Fund (Insolvency and Tail-Risk Backstop)

To protect workers if an employer fails or disappears, firms pay a small levy into a pooled Guarantee Fund. Our model shows that a levy of roughly 2% of the AI-driven profit uplift fully funds a 20% reserve buffer against the aggregate payout^[14]. The Fund also absorbs lifetime-payment liabilities for workers who never find comparable work, spreading that tail risk across all beneficiaries of automation rather than leaving it on a single firm’s balance sheet. This mirrors established wage-insurance and displacement-insurance designs, which have historically cost only a few dollars per worker per year to administer^[15].

6.4 Re-employment Top-Ups to Preserve Work Incentives

To ensure the scheme never discourages work, a worker who takes a lower-paid job receives a partial top-up (e.g. 50% of the wage gap, tapering over two years), so total income from working always exceeds the Continuity Dividend alone. This borrows directly from wage-insurance evidence, which shows such top-ups raise re-employment rates by 8–17 percentage points and can pay for themselves through restored tax receipts.

6.5 Mandatory Retraining and Portability

Continuity payments are conditioned on enrolment in accredited retraining where suitable, and the obligation is portable: it follows the worker, not the employer, and survives corporate restructuring, merger, or sale.

6.6 Tax Treatment and Anti-Avoidance

The government should make Continuity Dividend payments tax-deductible for the firm (as a legitimate cost of doing business) while taxing them as ordinary income for the worker. Strong anti-avoidance rules must prevent firms from reclassifying AI redundancies as voluntary resignations, performance dismissals, or offshoring to escape the obligation.

  7. Why Corporations Would Embrace It

A scheme that survives only through coercion will be fought and evaded. The Continuity Dividend is designed to be something corporations actively prefer to the alternatives. Six reasons stand out.

• It is cheaper than the alternatives. The likeliest substitute is a broad robot tax or AI levy to fund UBI—an open-ended charge on revenue or capital. The Continuity Dividend caps a firm’s exposure to its own displaced workers and ends when they are re-employed.
• It still leaves them far ahead. As Section 5 proves, firms retain ~40% of the original labour cost as net gain. Automation remains highly profitable; the model simply shares a slice of a large windfall.
• It protects their own customer base. Mass uncompensated displacement destroys the consumer demand that corporations depend on. Maintaining 75% of displaced income preserves aggregate spending — a collective-action problem the scheme solves industry-wide.
• It is a powerful ESG and reputational asset. “We automate responsibly — no worker we displace loses their livelihood” is a defensible social licence to deploy AI aggressively, reducing union, regulatory, and political friction.
• It accelerates, rather than blocks, automation. By defusing workforce resistance and the political backlash that could otherwise trigger punitive regulation, the model lets firms deploy AI faster and with certainty over their liabilities.
• The "dividend" framing aligns incentives. Positioned as sharing a productivity windfall rather than paying a fine, it lands far better with boards, employees, and the public — and converts a contentious cost into a competitive employer-brand advantage.

In short, the Continuity Dividend converts the single greatest social risk of corporate AI adoption — mass displacement — into a managed, capped, deductible, reputation-enhancing cost that still leaves automation overwhelmingly profitable.

  8. Conclusion

Universal Basic Income answers a real question — how do people live when machines do the work? — with the wrong instrument for AI-driven redundancy. It is universal in targeting the harm, taxpayer-funded when the gains are corporate, and unconditional when the goal is re-employment.

The Continuity Dividend corrects all three flaws. It is targeted, polluter-funded, and pro-work. The arithmetic is decisive: paying displaced workers 75% of salary still leaves firms roughly 40% ahead on every automated role, costs the economy about 0.34% of GDP, and is some eight times cheaper than universal UBI. With six government mechanisms — a legal definition, a Continuity Authority, a Guarantee Fund, re-employment top-ups, retraining and portability, and clear tax treatment — the model is enforceable and durable.

It is, finally, a model that corporations would choose. It caps their liability, protects their customers, strengthens their licence to automate, and reframes a feared cost as a shared dividend. Kinetic Consulting recommends the Continuity Dividend as a workable, financially proven, and politically resilient alternative to UBI for the age of artificial intelligence.

End

 

This white paper is provided for strategic discussion. Figures are illustrative models based on cited public benchmarks and should be calibrated to jurisdiction and sector before implementation.

Citations

[1]CNBC, “AI impacting labor market ‘like a tsunami’,” 2026. https://www.cnbc.com/2026/01/20/ai-impacting-labor-market-like-a-tsunami-as-layoff-fears-mount.html

[2]Goldman Sachs, “How Will AI Affect the Global Workforce?” 2025. https://www.goldmansachs.com/insights/articles/how-will-ai-affect-the-global-workforce

[3]The Guardian, “Universal basic income could soften hit from AI job losses, minister says,” 2026. https://www.theguardian.com/technology/2026/jan/29/universal-basic-income-used-cover-ai-job-losses-minister-says

[4] ShiftFlow / U.S. Bureau of Labor Statistics, fully-loaded labour cost (~42% burden), 2026. https://www.shiftflow.app/blog/labor-cost

[5]Britannica, “Universal basic income: Pros and cons of free money for all.” https://www.britannica.com/money/universal-basic-income-ubi

[6]Chiang Mai University, “History of the Universal Basic Income.” https://archive.lib.cmu.ac.th/full/T/2016/phil61216brp_ch4.pdf

[7]American Medical Association, “Universal Basic Income Pilot Studies (CMS Report 3),” 2021. https://www.ama-assn.org/system/files/2021-06/j21-cms-report-3.pdf

[8]Britannica, on Andrew Yang’s “Freedom Dividend.” https://www.britannica.com/money/universal-basic-income-ubi

[9]Ivàlua, “Review of Evidence: Universal Basic Income Pilot Project,” 2023. https://ivalua.cat/sites/default/files/2023-03/Literature_review_iva%CC%80lua_final_ANG.pdf

[10]McKinsey Global Institute, “AI could increase corporate profits by $4.4 trillion a year,” 2023. https://www.mckinsey.com/mgi/media-center/ai-could-increase-corporate-profits-by-4-trillion-a-year-according-to-new-research

[11]ShiftFlow / U.S. Bureau of Labor Statistics, fully-loaded labour cost (~42% burden), 2026. https://www.shiftflow.app/blog/labor-cost

[12]CEPR / NBER, “Wage insurance for displaced workers,” 2024. https://cepr.org/voxeu/columns/wage-insurance-trade-displaced-workers-middle-ground-alternative-rising-protectionism

[13]Goldman Sachs, “How Will AI Affect the Global Workforce?” 2025. https://www.goldmansachs.com/insights/articles/how-will-ai-affect-the-global-workforc

[14]McKinsey Global Institute, “AI could increase corporate profits by $4.4 trillion a year,” 2023. https://www.mckinsey.com/mgi/media-center/ai-could-increase-corporate-profits-by-4-trillion-a-year-according-to-new-research

[15]Brookings, “Insuring America’s Workers in a New Era of Offshoring,” 2016. https://www.brookings.edu/articles/insuring-americas-workers-in-a-new-era-of-offshoring/