Investors Lack Tools to Judge Trustworthy AI, GPAI Expert Group Says

A gap between policy and investment decisions

The authors, Celine Caira, Laurie Fitzjohn-Sykes, Paul Fehlinger and Sophie Walker, point to the scale of money flowing into the sector: in 2025, 61% of all venture capital funding was directed towards AI. Investors therefore influence which AI companies are funded and which business models succeed.

Policymakers have spent years building principles for trustworthy AI. The OECD AI Principles, described as the first intergovernmental standard on AI, were adopted in 2019 and revised in 2024. Yet according to the post, investors have no common way to check whether AI companies live up to them. Investment committees often lack the shared language, data, tools, experience and signals needed to turn trustworthiness into something they can evaluate and act on.

The post stresses that the disclaimer applies: the views are those of the authors and do not necessarily reflect the official views of the OECD, the GPAI or their member countries.

Trustworthiness seen as financially material

At the 11 June 2026 meeting, discussions across public and private markets produced one clear message: investors no longer view trustworthy AI as a purely ethical consideration. The central question for many is whether a factor affects long-term value creation, and increasingly they believe AI trustworthiness does.

• Public markets: responsible AI practices are seen as material through enterprise customer demand, regulatory exposure, insurability, liability risk, compute cost, reputational risk, AI incidents and cybersecurity failures.
• Private equity: responsible AI is becoming part of due diligence, Board engagement, portfolio management and preparation for future public listings.
• Venture capital: most firms see commercial opportunities in trustworthy AI, according to recent investor surveys, but many have yet to establish a shared understanding of what it means in practice.

The post also cites research by the Oxford Martin AI Governance Initiative, which it says shows that companies with strong governance and robust AI risk management may enjoy competitive advantages as customers become more selective.

Three things investors say they need

The authors argue that the biggest gap is information, not motivation. Three needs came up repeatedly across investment communities.

• Clearer frameworks. Resources such as the OECD AI Principles and the OECD Due Diligence Guidance for Responsible AI offer foundations, but investors often need simpler, more actionable guidance tailored to business models and investor-specific language.
• Comparable data. Investors want to know how companies govern AI systems, manage risks, conduct testing, oversee third-party providers and address incidents. Much of this information is currently unavailable, inconsistent or hard to compare. The Hiroshima AI Process (HAIP) Reporting Framework, the first global voluntary framework for structured and comparable reporting on AI governance and risk management, could help lay the foundation for more consistent assessments.
• Credible verification. Independent audits, third-party assessments and assurance mechanisms can help close the gap between company claims and investor confidence, mirroring developments in cybersecurity, sustainability reporting and financial auditing.

Without these building blocks, the post says, investment teams will keep struggling to treat trustworthy AI as a material factor in standard processes.

Opportunity, not only risk

The authors say investor groups agreed that too much of the current debate focuses only on risk. Companies with strong AI governance may gain customer trust more quickly, win contracts in highly regulated sectors and face fewer operational disruptions and lower compliance and liability costs. They may also be better positioned to scale sustainably as AI becomes embedded across the economy.

Several companies already treat AI governance as a business enabler rather than simply a compliance requirement, the post says, using it to strengthen customer relationships, support product development and stand out in crowded markets.

The GPAI Expert Group on Trustworthy AI Investment aims to bridge the gap by acting as a forum for policymakers, investors and AI experts, looking for ways to align investment incentives with standards such as the OECD AI Principles, the Hiroshima AI Process Reporting Framework and the OECD Due Diligence Guidance for Responsible AI. For the market, the message is that interest is growing, and the next step is turning it into practical, standardised guidance that can steer capital towards firms building and deploying trustworthy AI.