“AI is more suited to ‘move slow to move fast’ than ‘move fast and break things’.“
A New York-based partner at a billion-dollar-plus generalist VC firm made this comment as part of our research into how VCs are engaging with responsible AI. It’s striking coming from an industry built on disruption, and captures a shift in thinking that our data shows is far more widespread than popular narratives suggest.
Over the past 18 months, Reframe Venture, ImpactVC, and Project Liberty Institute have been running an extended research, interview, survey and engagement project on responsible AI in venture capital, involving over 200 VC and growth equity funds and 80 LPs managing more than $6 trillion in AUM across events in the USA, Paris, Berlin, London, Tokyo, and Singapore.
In our latest research, an in-depth survey of 56 venture capital professionals across the world, we found a surprising and concerning result: there is an overwhelming conviction that responsible AI is financially material, but VCs are struggling to integrate it into their decision making.
The finding was reinforced at our March 2026 responsible AI conference, which brought together 70 VCs and LPs in New York.
The alpha in responsible AI
Nearly three-quarters of the VCs we surveyed believe that companies with stronger responsible AI practices will be more financially successful — rising to 83% among those with more than five years of experience. And 84% see direct investment opportunity in companies that make responsibility core to their proposition.
The logic is straightforward: responsible AI avoids incidents and gains customers — high-profile collapses and lawsuits over AI-related harms have made the financial materiality of AI risk impossible to ignore. In turn, enterprise AI adoption is failing to live up to the predictions of AI zealots.
AI systems are massively improving in capability, but still lack the security, accountability, and transparency that businesses require before integrating them into real workflows. In addition, AI’s impact on human agency is a real, tangible operational challenge and social risk. Some 84% of surveyed VCs expressed concern about AI’s potential for disempowering its human users.
Against that backdrop, more stringent procurement expectations from large companies are turning responsible AI from a nice-to-have into a sales requirement. Reliable systems, transparent data handling, and human-augmenting designs are fast becoming the moat for real-world adoption. In regulated industries like healthcare, defense, and financial services, trust is a prerequisite. Across the board, the companies building trustworthy AI are the ones finding commercial success.
Building the ‘AI responsibility stack’
The current landscape — with Big Tech creating standards that mainly serve competitive interests, while nonprofits focus on frontier risks rather than deployment guidance — leaves a vacuum in practical, commercially viable safety infrastructure. This vacuum is an investable gap.
Some startups are capitalising on this market movement, with a booming sub-sector emerging in AI safety, assurance, and agency infrastructure.
At our recent New York conference, the business case for such innovations was made clear on stage by Zoe Weinberg, partner at ex/ante, a fund with an investment thesis built on turning AI risk into opportunity.
It’s not niche funds discovering this market gap: 91% of those surveyed see significant opportunity in the ‘responsibility stack’, drawing a clear parallel to the emergence of cybersecurity over the past decade.
LP pressure
Despite this conviction, only 14% of VCs surveyed rated their own firm’s AI risk assessment capabilities as “good.” And just 27% feel they have sufficient internal expertise — a dramatic contrast to the 74% who feel confident on ESG generally.
Regulation hasn’t helped here: while raising the topic’s importance, uncertainty about requirements, scope, and timelines has become a persistent frustration, even more so than the actual compliance costs. This has made it difficult for firms to commit resources against a moving target.
VCs are starting to develop their thinking, expertise, and processes to overcome their knowledge gaps — with a push from their LPs.
In the US, where federal AI regulation remains fragmented and ESG-specific language carries political baggage, LPs appear to be stepping into a governance vacuum — asking direct questions about AI risk management in ways that bypass the ESG framing entirely. European funds have been subject to regulations longer, and therefore report less LP pressure.
Still, about half of VC respondents, regardless of geography, expect responsible AI to come up in their next fund raise.
The upstream pressure is beginning to shift incentives, budgets, and interest in learning more about the topic. Reframe Venture resources such as our Responsible AI Due Diligence Tool and Responsible AI training are seeing fast-growing interest from VCs across the world.
There is a strong and widely held belief among VCs globally that responsible AI is good business. The data supports it, the market is demanding it, and the investment opportunity is real. But belief without implementation is just talk — VCs must step up to realize the commercial opportunity, and LPs are likely to reward them for it.
_________________________________________________________________________________________
Oliver Nixon is research lead at ReFrame Venture.