The investor's case, in four lines
Ethical due diligence is not a values exercise. For an investment group, it is a discipline that delivers four things at once.
Surface the ethical, human rights, and AI failure modes that hit valuations, trigger regulatory action, or cost a company a key customer, before they reach the deal memo or the board pack.
Give LPs, customers, regulators, and talent a credible, evidence based reason to believe what your portfolio companies say about their data and AI practices.
Turn ethical posture into a procurement advantage, a hiring advantage, and a market access advantage in jurisdictions that increasingly reward it.
Stop value leaking through avoidable incidents, forced product withdrawals, regulatory remediation costs, and the dilution that follows an unplanned crisis raise.
Each of these is reachable through the same underlying work. The rest of this article is how that work is structured, and who is already doing it.
A working definition
Ethical due diligence is the structured assessment of how a company creates, uses, and governs data and AI systems, the ethical and human rights effects those systems produce, and the company's capacity to keep those effects under control as it scales.
For an investor, it answers a single question. Will this asset compound trust, or quietly burn it?
Why investors are moving on this now
Three forces converge.
Regulation has teeth. The EU AI Act, GDPR enforcement, the Digital Services Act, and sector specific rules in finance, health, and HR. Non compliance now carries fines, market access risk, and personal liability for executives. The EU Corporate Sustainability Reporting Directive goes further, requiring large institutional investors to assess and disclose human rights risks across their portfolios, including impacts that are severe and likely, not only those that are financially material.
LP expectations have shifted. Limited partners increasingly ask for evidence, not assertions, on responsible technology practices across the portfolio. The Principles for Responsible Investment now count roughly 5,000 signatories representing about $128 trillion in assets, and PRI's 2025 Awards explicitly recognised ethical AI work in the Human Rights category.
The numbers favour ethical operators. A March 2026 survey by ReframeVenture, Project Liberty Institute, and ImpactVC, drawing on LPs and GPs managing over $500 billion, found that more than 90% see investable opportunity in responsible AI infrastructure, and that 73% of investors believe firms with stronger responsible data and AI practices are more likely to succeed financially, rising to 83% among investors with five or more years of experience.
A small but growing cohort of investment groups has noticed the asymmetry. Ethical operators tend to be undervalued at entry and revalued upward as the regulatory and reputational gap widens. Backing them deliberately is a strategy, not a slogan.
This is already happening, in the open
The trend is not theoretical. Several named, dated, verifiable moves anchor it.
- A pension fund and a global institution wrote the playbook. In June 2024, the World Economic Forum and Canada Pension Plan Investments published the Responsible AI Playbook for Investors. It tells investors directly to conduct Responsible AI due diligence on the portfolio and to engage external managers and the broader ecosystem on the same terms. CPP Investments had already updated its 2023 Policy on Sustainable Investing to include responsible sourcing and use of AI as a recognised sustainability factor.
- The first VC specific toolkit shipped in December 2025. Project Liberty Institute, ReframeVenture (a network of more than 150 LPs and 500 VCs), and ImpactVC (700+ members) released the Responsible AI Due Diligence Toolkit, built specifically to plug into existing diligence flows with sector and stage sensitive questions.
- The profession's standards body has a framework. The CFA Institute publishes Ethics and Artificial Intelligence in Investment Management, and follow up commentary makes the case that ethical decision frameworks are now critical infrastructure for AI in capital markets. IC memos can reference it by name.
- European funds are building theses around it. AI.FUND is one of several European VCs explicitly backing AI champions that are globally competitive, compliant with the EU AI Act, and anchored in European values. The combination of regulatory readiness and ethical posture is increasingly part of the pitch to LPs, not separate from it.
- The institutional groundwork is older than people remember. The British Institute of International and Comparative Law and PRI convened private equity firms on human rights due diligence as far back as 2017. The vocabulary has been ready for a decade. What has changed is that AI gave it new urgency and that regulation gave it teeth.
The signal is consistent. Pension funds, VC networks, regulators, and the profession's own standards body are converging on the same answer: ethical and human rights diligence over data and AI is now part of the work.
What ethical diligence actually examines
ESG screening covers the broad terrain of environment, social, and governance. Ethical diligence is narrower and sharper. It looks at how a company's products and operations affect people, and how robustly those effects are managed.
In practice, the questions are:
- Data lineage. Where does training and operational data come from, on what legal basis, and whose interests does it represent.
- AI systems in production. What decisions do they influence, with what human oversight, and what is the realistic failure mode.
- Human rights footprint. Who is affected by these systems, and have those groups been considered in design, testing, and monitoring.
- Governance scaffolding. Named owners, escalation paths, model and data documentation, audit trail.
- Cultural signal. Does leadership treat ethics as a constraint to minimise, or as a property to demonstrate.
Each is a question with an evidence based answer. None is captured by a generic ESG questionnaire.
Curious how your portfolio answers these questions?
Let's examine it together ↗The three layers of an ethical diligence engagement
A well run programme works on three levels in parallel.
The first deliverable is a clear picture: where does this company sit today on data ethics, AI governance, and human rights impact. Strengths, gaps, and the specific risks that would matter at the next funding round, exit, or regulatory event. Without that map, every later recommendation is opinion.
For existing holdings, diligence becomes a quarterly or annual rhythm that surfaces drift before it becomes a headline. For prospective deals, it sits alongside legal, financial, and technical diligence and feeds directly into the term sheet and the post close plan. The output is a prioritised set of capacity building activities the company can absorb without losing momentum.
Reports that no one acts on are wasted budget. The third layer is targeted advisory and training so founders, product leaders, and engineers can run the controls themselves. The investor does not become the ethics function. The portfolio company becomes capable of being its own.
Where founders quietly resist, and how to handle it
Even motivated founders push back on ethical diligence in predictable ways.
- "We are too early for this." Most controls are cheaper to install at 30 people than at 300. Founders who treat governance as a late stage problem inherit a much harder version of it.
- "Our customers don't ask about it." They will. Enterprise procurement, public sector, and regulated industries already do, and the rest are catching up. The market signal is in the ReframeVenture data already cited: more than 90% of investors see opportunity in the responsible AI stack precisely because customers are starting to ask.
- "We cannot afford to slow down." A good engagement does not slow the roadmap. It removes the risks that would have stopped the roadmap entirely.
- "This is a cost centre." It is, until the first competitive bid where ethical posture is the tiebreaker, or the first incident that does not happen.
The investor's role is to make these conversations routine, not adversarial.
Signals of an investable ethical posture
Across many conversations with founding teams, a few signals correlate with companies that handle this well at scale:
- A named senior accountable owner for AI and data ethics, not buried under legal or IT.
- Documentation that is read, not just produced. Model cards, DPIAs, and incident logs that show signs of actual use.
- A willingness to say no to a feature, a data source, or a customer when the ethical case is weak.
- Engineers who can describe the failure modes of their own systems without prompting.
- A leadership team that does not flinch when asked uncomfortable questions about a deployed product.
Absence of these signals is not a disqualifier. It is a price discovery moment.
A staged playbook for an investment group
For firms starting from scratch, three steps are enough to install the discipline. Each one builds on the last and can move at the pace the firm chooses. The structure deliberately mirrors the WEF and CPP playbook so the work plugs into a recognised framework rather than reinventing one.
- Step 1. Establish the lens. Agree on the ethical and human rights criteria the firm cares about. Borrow what works from the WEF playbook and the Project Liberty toolkit. Translate the result into a concise diligence checklist that fits inside the existing investment process.
- Step 2. Baseline the portfolio. Run a lightweight ethical map across current holdings. Triage by exposure and by deal stage. Identify the two or three companies where the next intervention will create the most value.
- Step 3. Operationalise. Set the rhythm. Regular review for high exposure assets, lighter touch for the rest. Add the ethical diligence pack to the standard new deal workflow. Brief LPs on the change, and reference the WEF, CFA Institute, and PRI material to anchor the conversation in shared vocabulary.
Once the three steps are in place, ethical diligence stops being a side project. It becomes part of how the firm picks, prices, and supports companies.
Want a version of this sequence shaped to your firm?
Let's plan your own ↗Frequently asked questions
Is this an established practice or an emerging one?
It is established at the top end and emerging across the rest of the market. CPP Investments and the World Economic Forum have published a joint playbook. The CFA Institute has issued a profession wide framework. A VC specific toolkit shipped in December 2025. The market is past the pilot phase.
Is ethical diligence the same as ESG screening?
No. ESG screening is broad and standardised. Ethical diligence is targeted at the specific harms a given company's products and operations can produce, especially through data and AI.
Does this only matter for AI heavy companies?
No. Any company that processes personal data, automates decisions, or deploys models inherits ethical exposure. AI native companies carry it more visibly, not exclusively.
Will this slow our deal flow?
Not when it is integrated into the existing process rather than bolted on. The diligence work itself takes days, not weeks, once the framework is in place.
What is the deliverable for the investment committee?
A short ethical diligence memo: posture, top three risks, recommended remediations, and an assessment of whether the founding team is the right one to execute them. Reference the CFA Institute framework where useful so the IC has a shared standard to evaluate against.
References
- World Economic Forum and CPP Investments Insights Institute. Responsible AI Playbook for Investors, June 2024. weforum.org.
- CPP Investments. Why every investor should embrace Responsible AI. cppinvestments.com.
- Project Liberty Institute, ReframeVenture, ImpactVC. Responsible AI Due Diligence Toolkit, 22 December 2025. projectliberty.io.
- ReframeVenture, Project Liberty Institute, ImpactVC. Pioneering VC Survey: Nine in Ten Investors See Financial Opportunity in a Responsible AI Stack, 18 March 2026. projectliberty.io.
- CFA Institute. Ethics and Artificial Intelligence in Investment Management: A Framework for Professionals, 2022. rpc.cfainstitute.org.
- CFA Institute. Why ethical decision frameworks are critical for AI in investment management. cfainstitute.org.
- Charles Russell Speechlys. Growing investor focus on human rights and social sustainability, 2026. charlesrussellspeechlys.com.
- Principles for Responsible Investment. unpri.org.
- British Institute of International and Comparative Law and PRI. Human Rights in Private Equity Investment. biicl.org.
- AI.FUND. ai-fund.vc.
Independent data ethics and responsible AI consulting. EU Commission Independent Expert on AI, ODI-certified, qualified DPO, and lecturer at leading European business schools. Learn more →