A framework for investors who refuse to profit from armed conflict – with an intent‑based trigger and a sanctions failsafe
1. The Investor’s Dilemma
You are a GP. Your fund backed a satellite imaging startup built for flood monitoring. Then a war broke out. The startup’s revenue tripled – partly from humanitarian agencies tracking refugees, partly from a defence contractor using the same imagery for targeting. Your LPs ask: are we making money from bombs?
You cannot avoid dual‑use exposure. But you can avoid war windfall – the incremental profit directly attributable to supplying destructive capabilities during an active conflict. The challenge is drawing a line between legitimate civilian business in war zones (food, medicine, refugee comms) and profit from causing harm.
This framework uses a single, principled yardstick: the intended function of the contract. If the purpose is to cause harm or kill, the policy activates. If the purpose is to reduce harm, heal, or is neutral, the policy does not activate – even if the customer is a military.
In addition, a standalone sanctions trigger captures cases where a portfolio company is formally designated as a threat by credible authorities, regardless of contractual intent.
2. Baseline Condition: Conflict Escalation
No policy activates without a clear, externally verifiable escalation to armed conflict – except under the standalone sanctions trigger (Section 4). Acceptable sources:
- UN Security Council resolution or statement recognising a state of war.
- Formal declaration of war by a recognised state.
- European Union or US government determination of “active hostilities” in a region where a portfolio company operates or sells.
Without conflict escalation, the primary pathway (Triggers 1‑3) does not activate. Peacetime defence contracting is outside that pathway.
3. Monitoring Signals (No Activation)
Some events raise questions but do not automatically trigger neutralisation. They trigger a monitoring report to LPs within 60 days.
| Signal | Action |
|---|---|
| Valuation surge | Portfolio company valuation increases by >30% within a conflict period, uncorrelated with civilian market growth. GP commissions a review to determine whether the surge is driven by contracts intended to cause harm, or by legitimate civilian demand. |
| Revenue from conflict region | Portfolio company revenue from a conflict‑affected region exceeds 15% of total revenue, and that revenue includes sales to belligerent‑state entities. GP investigates the intended function of those sales. |
If the investigation shows no contracts intended to cause harm, the policy does not activate. LPs receive a confidential report.
4. Activation Triggers (Neutralisation Required)
The policy activates when either Pathway A or Pathway B is satisfied.
Pathway A – Intent‑Based Defence Contract (all three conditions must be met)
- Conflict escalation (Section 2) is ongoing.
- The portfolio company has a defence contract with a clearly military end‑user (state military, intelligence agency, or prime defence contractor).
- The contract’s intended function is to cause harm or kill in direct support of combat operations.
Examples of harm‑causing functions (non‑exhaustive):
- Targeting, munitions guidance, fire control, battle management for frontline combat.
- Weapons or ammunition components.
- Sensor or intelligence data used for lethal targeting.
- Logistics software explicitly designated for moving weapons or troops in direct support of combat.
Excluded from activation (even with a defence contract):
- Contracts intended to reduce harm, prevent deaths, or heal – medical evacuation, field hospitals, trauma care, demining, refugee tracking, disaster response, humanitarian logistics.
- Contracts that are neutral – general communications infrastructure, food/water supply, shelter, energy systems, unless customised for a harm‑causing purpose.
Pathway B – Sanctions or Proscribed Support (standalone)
The portfolio company:
- Appears on any official sanctions list published by the European Union, the United Kingdom, Canada, Australia, or an East African Community Partner State; or
- Is demonstrated, through credible evidence (e.g., UN reports, judicial findings, or official government designations), to be in direct support of a proscribed organisation or paramilitary group engaged in armed conflict.
If Pathway B is met, the policy activates immediately, regardless of conflict escalation or contractual intent. This provides an objective failsafe for clear‑cut cases of malign activity.
5. The Three Policy Options
Once activated, the GP must choose one of the following actions:
| Option | Description |
|---|---|
| A – Full Divestment | Sell the entire stake. The war‑windfall portion of the gain is ring‑fenced and donated to humanitarian causes. |
| B – Segregated Accounting | Retain the stake but ring‑fence the war‑windfall portion of unrealised/realised gains. That portion is never distributed to LPs; it is held in escrow and donated when realised. |
| C – Proxy Activism | Use board rights to force the portfolio company to adopt its own war‑windfall reduction mechanism. If the company refuses within 12 months, divest under Option A. |
6. War‑Windfall Calculation
The war windfall is the excess gain attributable to the harmful activity (Pathway A or Pathway B), not to the company’s baseline civilian business.
- Baseline: valuation or revenue growth projected from civilian markets, using the 24 months before the trigger event (conflict escalation for Pathway A, or the date of sanctions listing / evidence confirmation for Pathway B) as a reference.
- Excess: actual growth minus baseline, adjusted for any identifiable non‑harmful contracts.
- Attribution: only the portion of excess directly linked to the harmful activity is neutralised. If attribution is impossible, the GP neutralises a conservative estimate (e.g., 50% of excess) and documents the methodology.
A third‑party valuation specialist certifies the calculation.
7. LP Communication & Consent
| Stage | Action |
|---|---|
| Initial disclosure | LPA or side letter includes the full policy, including both pathways. |
| Monitoring report | Within 60 days of a monitoring signal, LPs receive a confidential analysis. |
| Activation notification | Within 30 days of an activation trigger (Pathway A or B), the GP sends a report detailing the harmful activity, estimated war windfall, and proposed action. |
| LP consent | For material actions (divestment, reallocation), GP seeks LP consent unless pre‑authorised. |
| Annual transparency report | Summary of monitoring signals, activation events, and neutralisation disbursements. |
8. Neutralisation – Where the Money Goes
Ring‑fenced war windfall is donated to pre‑approved humanitarian organisations with no military affiliation:
- International Committee of the Red Cross (ICRC)
- UN High Commissioner for Refugees (UNHCR)
- Médecins Sans Frontières (MSF)
- Demining and unexploded‑ordnance clearance organisations
- Refugee support and trauma counselling
LPs may nominate additional charities, subject to GP approval.
9. Fallback – When Separation Is Impossible
If the war windfall cannot be reliably isolated, the GP may:
- Neutralise all returns from the trigger period for the specific portfolio company, up to a cap of 2× the original investment. This cap is subject to update.
- Offer LPs a conscience exit – withdraw capital without penalty – and donate the corresponding GP carry.
10. Relationship to the Founder Framework
This investor protocol mirrors the previous founder‑level framework. Together, they create a clean chain from conflict‑derived revenue to humanitarian neutralisation – without over‑reacting to legitimate civilian business in war zones.
11. Conclusion: Intent Is the Yardstick, Sanctions Are the Failsafe
The refined policy uses a dual‑path structure:
- Pathway A asks: Is the contract’s intended function to cause harm or kill, during an active conflict?
- Pathway B asks: Is the company officially sanctioned or directly supporting a proscribed armed group?
If either answers yes, the policy activates. The profit is ring‑fenced and donated to humanitarian causes. If the answer is no (harm reduction, neutral, or no sanctions), the policy does not activate – even in a conflict zone, even with a military customer.
This is neither naive nor paranoid. It distinguishes a startup that helps field hospitals from one that helps targeting computers, and it provides an objective failsafe for clear‑cut cases of malign activity. GPs and LPs gain a clear, principled rulebook – and a process for the grey zones.
The question is not whether you will ever face a war windfall. The question is: have you built a framework that can tell the difference between healing and harming, and neutralise only the latter?
This blog post is part of a series on governance models for dual‑use technology. The framework described is under active development and is not legal advice