The Defensibility Question: Helen Fan on What Survives When Frontier Models Climb the Stack

The Defensibility Question: Helen Fan on What Survives When Frontier Models Climb the Stack

🎙️ This week we sit down with Helen Fan [https://www.linkedin.com/in/helenfanlegalai/] — California lawyer, Chief AI Officer at a Silicon Valley boutique firm, and one of the most original voices currently writing and building in legal AI. What makes Helen unusual is the combination. She is a practising cross-border lawyer who is also genuinely technically literate — building, testing and writing about agentic systems at a level most lawyers (and a fair few engineers) do not reach. And she sits squarely on the bridge between the US and China technology frontiers (building legal tech communities in each). "My role is really about being a bridge — between tech people and legal professionals, and between the US and Asia." It is a rare vantage point — and one that has produced one of the more original frameworks we have come across in the legal AI conversation this year. Our conversation is in two halves. First, we ground the conversation in the Legal AI Value Stack — Helen's five-level framework for thinking about defensibility in a world where frontier models keep climbing the stack. Then we get into the fun part: OpenClaw LLP, her public experiment to actually walk the roadmap herself. We get into the design choices behind her two AI associates, the four-layer security framework she has built, the Argument Report skill, her custom debrief skill which has her agents look back over recent work and recommend tweaks, why she ended up on Discord, and her honest reflections after the first 60 days. Key References Connect with Helen Fan — California lawyer & Chief AI Officer | LinkedIn [https://www.linkedin.com/in/helenfanlegalai/] | Substack [https://helenfan1.substack.com/] | Personal site [https://helenlab.com/] The Legal AI Value Stack — Five Levels of Defensibility [https://helenfan1.substack.com/p/the-legal-ai-value-stack-five-levels] — Helen's now widely-shared framework. Start here if you have not read it yet. The OpenClaw LLP AI-Native Law Firm Experiment [https://helenlab.com/openclawlaw/] — Helen's public 100-day project, documented on LinkedIn and her site. Helen's Stanford Presentation [https://www.youtube.com/watch?v=ZG1Js1LT86I] — A quick video run-through of the talk Helen gave at Stanford a few weeks before this recording. If you enjoyed this conversation, please share it with someone or a community who you think would find it valuable . And if you have a moment, rate the show and tell us what landed — it helps us reach more people and keep getting brilliant guests like Helen. --- For more conversations at the intersection of law and technology, head to⁠⁠⁠⁠⁠ https://lawwhatsnext.substack.com/⁠⁠⁠⁠⁠ [https://lawwhatsnext.substack.com/].

Legal Engineering's Moment - Mary O'Carroll on her new CEO Role and the skill set that is becoming highly prized

🎙️ This week we sit down with Mary O'Carroll [https://www.linkedin.com/in/maryshenocarroll/] - the newly announced CEO of Legal Eng Consulting Group (LECG) [https://www.lecg.com/], and one of the most influential figures in the legal operations and legal technology industry. Mary's career arc is, by her own admission, a “portfolio” one. She built and led legal operations at Google. She founded and grew the CLOC community [https://cloc.org/] into the global home of legal ops. She moved into big law as Chief Operating Officer at Goodwin [https://www.goodwinlaw.com/en]. She has sold legal technology (Ironclad [https://ironcladapp.com/]), advises investors (she is an Executive in Residence at Signal Fire [https://www.signalfire.com/]), technology companies (Sandstone [https://sandstone.com/]), hosts her own podcast Pearls On Gloves Off [https://open.spotify.com/show/0DrT5PPH1gKOJQWkygHPV9], authors her own Newsletter [https://maryocarroll.substack.com/], and — as of yesterday is now the full-time CEO of LECG, a boutique legal operations consulting and services firm focused on what she calls legal engineering.  --- Alex and I are big fans and friends of Mary - few people in the legal industry benefit from her wide ranging experience and perspective. Even fewer can so succinctly distill the objectives, incentives, opportunities and competing priorities of the firms, teams, tech companies and individuals that make up our unique legal ecosystem, and none with the energy, candour and piercing intellect Mary brings to every conversation.   --- In the short time we spend together we get into: * The CEO Announcement — Mary uses the episode to confirm she is joining LECG as full-time CEO. She explains why!  * What a Legal Engineer actually is? Mary defines the legal engineer as someone who can understand legal processes, re-engineer them, and then leverage technology to automate, accelerate, optimise and maintain them.  * The Billable Hour is the Operating System — Drawing on her recent Substack piece, Mary explains why everyone has been focused on the wrong half of the billable hour problem. Pricing on a fixed fee is the easy bit. The harder bit is the underlying incentive structure of a law firm.  If you give an associate a tool that turns eight hours into one and tell them they still need to hit hours, you have created a structure that is at odds with the technology itself. It is, she says, "a hundred percent the innovator's dilemma." * The View from the VC Side — As an Executive in Residence at Signal Fire, Mary has watched venture capital wake up to legal in a serious way.  --- Connect with Mary O'Carroll — through Legal Eng Consulting Group [https://www.lecg.com/about-us] (LECG) | on Substack [https://substack.com/@maryocarroll] | or LinkedIn [https://www.linkedin.com/in/maryshenocarroll/] --- If you enjoyed this conversation, please share it with someone or a community who you think would find it valuable . And if you have a moment, rate the show and tell us what landed — it helps us reach more people and keep getting brilliant guests like Mary. --- For more conversations at the intersection of law and technology, head to⁠⁠⁠⁠ https://lawwhatsnext.substack.com/⁠⁠⁠⁠ [https://lawwhatsnext.substack.com/].

The Quantum Paradox: Rebecca Keating and Laura Wright on the Race to Get Encryption-Ready

Google says it will be able to break RSA encryption by 2029. Third-party actors are already collecting encrypted data on the assumption they'll be able to read it later. The UK has just committed £2 billion to a quantum strategy. 🎙️This week we sit down with Rebecca Keating [https://www.linkedin.com/in/rebeccakeating/] and Laura Wright [https://www.linkedin.com/in/laura-wright-817a93116/] — barristers at 4 Pump Court [https://www.4pumpcourt.com/] and the co-authors of A Practical Guide to Quantum Computing and the Law [https://www.lawbriefpublishing.com/2024/12/free-chapter-from-a-practical-guide-to-quantum-computing-and-the-law/]. Both are a rare breed of Barrister with technical credentials to complement their deep legal expertise. Rebecca worked in-house at Dropbox before being called to the Bar in 2017, sits on the ICO's Technology Advisory Panel, and has acted in one of the only quantum-related cases to pass through the UK courts. Laura took an MSc in Computing Science at Imperial mid-career — her final project was a new coding language for legal contracts — and now writes and speaks regularly on smart contracts, AI liability, and quantum risk. --- What You'll Learn * What is Quantum Computing — Alex surprises us all with his own definition and a sneak preview into how he likes to prepare for our podcast conversations! 👀 * The Quantum Paradox — Rebecca's framing for the central tension of the technology - the ability of quantum computers to upend the security systems that are the basis upon which we keep information safe, but there's also the capability to have even more secure systems than we have ever had. * Harvest Now, Decrypt Later — This is not a future threat. Third-party actors are already collecting RSA-encrypted data they can't read today on the assumption they'll be able to decrypt it within a few years. NIST's quantum-readiness window of 2030–2035 is, in Rebecca's view, too late to start the conversation — particularly for anyone holding sensitive medical, political, or nationally significant data. * Contracting for Quantum Computing as a Service — Customers won't own quantum computers — they'll access them remotely on a pay-as-you-go basis. Laura walks through what features are likely to make "QCaaS" contracts genuinely different from SaaS. --- Connect with Rebecca Keating [https://www.linkedin.com/in/rebeccakeating/] — Barrister at 4 Pump Court [https://www.4pumpcourt.com/our-people/rebecca-keating/] | Member, ICO Technology Advisory Panel Connect with Laura Wright [https://www.linkedin.com/in/laura-wright-817a93116/]— Barrister at 4 Pump Court [https://www.4pumpcourt.com/our-people/laura-wright/] | Co-host, 4 Pump Court podcast Their book — A Practical Guide to Quantum Computing and the Law [https://www.lawbriefpublishing.com/2024/12/free-chapter-from-a-practical-guide-to-quantum-computing-and-the-law/] (Law Brief Publishing, December 2024). The Law of AI (2nd edition, Sweet & Maxwell) [https://www.amazon.com/Law-Artificial-Intelligence-2nd/dp/0414114507]— Rebecca and Laura author the chapter on AI and Professional Liability. Society for Computers and Law (SCL) [https://www.scl.org/about-scl/] — Rebecca and Laura's recent SCL webinar on quantum legal issues was the catalyst for this episode.  Both (+ Tom) are members of the SCL - a leading educational charity for the tech law community in the UK. --- If you enjoyed this conversation, please share it with someone or a community who you think would find it valuable . And if you have a moment, rate the show and tell us what landed — it helps us reach more people and keep getting brilliant guests like Rebecca and Laura. --- For more conversations at the intersection of law and technology, head to⁠⁠⁠ https://lawwhatsnext.substack.com/⁠⁠⁠ [https://lawwhatsnext.substack.com/].

Inside the Machine: Bilva Chandra on Trust, Truth, and the Future of Knowledge in an AI World

🎤 This week we sit down with Bilva Chandra [https://www.linkedin.com/in/bilva-chandra/] — who has spent the bulk of her career working on AI safety, ethics, and governance at the places where it matters most.  Her CV reads like a guided tour of the AI governance landscape: she's worked at OpenAI, RAND, the US AI Safety Institute (now CAISI), and most recently Google DeepMind. She's been inside the frontier labs building the technology and inside the institutions trying to govern or influence its development — often thinking about the same problems from both sides of the table. --- One of the catalysts for our conversation emanated from her recent contribution to a Google DeepMind paper — Architecting Trust in Artificial Epistemic Agents [https://arxiv.org/abs/2603.02960] — exploring what happens when AI systems become active participants in how knowledge is created and shared. But it's Bilva's broader career at the intersection of AI and society that makes this conversation so compelling: she's someone who genuinely cares about getting this right, and isn't afraid to say when she's worried. We cover a lot of ground — from the practical challenge of making AI systems reliable enough for enterprise adoption, to the deeper worry about what happens to human judgment when cognitive work is increasingly offloaded to machines. What emerges is a picture of someone who is both genuinely optimistic about what AI can unlock and deeply clear-eyed about the societal fault lines it's accelerating. Bilva doesn't treat AI risk as a theoretical exercise. She frames it as a human problem — one tangled up with polarisation, declining trust in institutions, and an information environment that was already broken before the first LLM shipped. --- Connect with Bilva Chandra — on LinkedIn [https://www.linkedin.com/in/bilva-chandra/] | Or by subscribing to Role Model [https://rolemodel.beehiiv.com/], her new newsletter on AI and society. --- If you enjoyed this conversation, please share it with someone who you think would find it valuable — especially anyone grappling with how to govern AI responsibly in their organisation. And if you have a moment, rate the show and tell us what landed — it helps us reach more people and keep getting brilliant guests like Bilva. --- For more conversations at the intersection of law and technology, head to⁠⁠ https://lawwhatsnext.substack.com/⁠⁠ [https://lawwhatsnext.substack.com/].

The AI Dividend: David Bushby on Corporate Teams Going It Alone

🎤 This week we sit down with David Bushby — Head of Legal Operations at Canva and the voice behind In Counsel Weekly [https://www.incounsel.com.au/incounsel-weekly] — for a conversation that's part provocation, part practical playbook, and part three mates who love this stuff catching up over a good topic.   David follows Sam Lewis as the second member of the brilliant Canva legal team to join us for a chat, and from our vantage it's easy to discern the talent, curiosity and sense of fun they must have coursing through their team. --- David has spent the last couple of years at the forefront of AI adoption in legal — not as a commentator, but as a practitioner. Custom GPTs, Claude skills in Cowork, vibe-coded legal research tools, Chrome extensions that make SaaS platforms do things they were never designed to do. He's materially contributing to the building of an AI-native legal function in real time. --- "I just started to feel, by the end of last year — we just really need to go it alone. We can control what tools we use... maybe the AI dividend is just up to us on the in-house side." Here's a big question David has been grappling with (and we get into it straight away during our catch up): are law firms passing on their AI productivity savings to clients? The invoice data says no. The rate increases — 12–16% in the US and UK, with individual partner hikes of 25–35% — say definitely not. And at the line-item level? Nothing.  So what does a smart in-house legal ops team do? They begin to contemplate going it alone.  David walks through Canva's owners mindset approach to outsourcing, the vibe-coded tools his team is building in Claude Code, and why the AI dividend might just be something in-house teams have to take for themselves. --- Connect with David Bushby — by subscribing to In Counsel Weekly [https://www.incounsel.com.au/incounsel-weekly] — David's popular bite-sized weekly newsletter for in-house counsel | Or find him on LinkedIn [https://www.linkedin.com/in/davidbushby/]  --- If you enjoyed our conversation, please do share it with someone who you think would be interested in listening (or is equally passionate or enraged by law firm productivity). And if you've got a spare minute, please rate the show and tell us what landed 🙏 It helps us grow our audience and continue to attract great guests like David. --- For more conversations at the intersection of law and technology, head to⁠ https://lawwhatsnext.substack.com/⁠ [https://lawwhatsnext.substack.com/].