How do you trust AI with your most sensitive data? That was the question that set the tone for Monday 20th April - Web3 & AI Dinner at the House of Lords. Co-hosted by Brian D. Evans, Sami R., Introduction.com and D2 Impact Group. 35 VCs and family offices across Web3, AI and frontier tech came together, alongside Lords, MPs and Katie Lewis 🚀 - a focused room where capital, technology and policy intersected in real time. Lord Taylor of Warwick opened the evening with a personal and inspiring reflection on the role of technology in shaping society - delivered with humour, clarity, and his signature use of props. Both entertaining and genuinely thought-provoking. We also heard from Katie Lewis 🚀 and Erika Ziolkowski, each bringing thoughtful perspectives on how AI and Web3 are evolving in practice. And as always, thank you to Dr Lisa Cameron - not only for her signature insights on Web3 and AI, but for setting the tone throughout the evening as an outstanding MC. Sami R. followed with a clear and timely message: AI is not limited by intelligence 0 it is limited by trust. The highest-value use cases sit behind sensitive data, where “trust us” is no longer enough. The shift ahead is from promise to proof - where privacy is enforced at the infrastructure level. Whoever solves trust in AI will define its real adoption. That is where the conversation moved. Not just what is being built, but what can actually be trusted, adopted, and deployed where it matters. Because the real question is not whether AI works. It’s whether the most sensitive institutions will ever be able to use it fully. Governments. Defence. Healthcare. Financial systems. Critical infrastructure. These are not edge cases, they are where the highest-value decisions sit. And today, they are still held back by one constraint: trust. So what actually unlocks adoption? Better models, or a fundamentally different trust architecture? And more importantly - what would it take for you to trust AI with your most sensitive data? Where do you see the real blocker today: capability, or trust? #AI #Web3 #DigitalAssets #FamilyOffices #VentureCapital #PrivateMarkets #Technology #HouseOfLords
-
-
-
-
-
+10
Interesting take. Personally, I’m comfortable using AI for sensitive work as long as there’s a clear human in the loop. AI can assist, but final judgment and control should stay with humans. Curious how others are balancing this in practice.
We work closely with clients who handle sensitive operational data. The conversations are always the same. The AI output looks promising. But the moment you ask where the data goes, who has access, and what the audit trail looks like, the room goes quiet. The shift from promise to proof is the right frame. Privacy enforced at the infrastructure level, not declared in a policy document. Whoever builds that layer credibly will unlock adoption in the places that actually matter. Donna Chen
amazing job Donna Chen! definitely a night to remember! 👏 thanks for organising such a great and inspiring event
We are seeing increasing demand for local LLMs that aren’t linked to or from the big labs - they might not have the power of the big labs but they’re close enough to cause local revolutionary evolution in governments and organisations all the data is safe, secure and kept within organisational walls. That’s the future that’s being written by teams like us, our colleagues and allies based on the overwhelming demand from governments, companies and all kinds of organisations. It’s worth thinking about - thanks to you all for the inspired posts and ideas !💡
The trust gap in AI is a compliance engineering challenge as much as a technology one. The shift from "trust us" to technical proof is exactly what financial institutions, and ultimately governments, need before they open their most sensitive data to AI systems. Did any of the Lords offer a concrete legislative angle on accelerating that shift?
🎯
AI adoption is no longer constrained by capability, but by trust, especially in high-stakes, data-sensitive environments. The next phase will be defined by infrastructure that proves security and privacy, not just promises it. Donna Chen
FURTENBURG supports you.
One for the books, already looking forward to the next!
James Bowater Bob Bubeck Alexander Fink Patrick Martin Chrissy Hill Jessica Maria Rota Imran Akram Jerry Wu Stanley Chan Amin B. Caroline Thomsen Rachel Vetch Lex Sokolin Jake Blew Ayt B. Theo Golden, CFA Shahid Qazi Kay Meyer Danish Chaudhry