Jack Newton

Just published my analysis on the legal industry's $900B repricing event - how AI is ending the billable hour and creating the biggest disruption in professional services history. Here is the full analysis: https://lnkd.in/gWXKEBbY While most focus on AI tools helping lawyers work faster, the real revolution is AI-native law firms replacing the entire business model. BigLaw convinced clients that time spent = value delivered, creating the only major industry where efficiency threatens profitability. That protection is about to expire. We're witnessing a fork that will split the legal landscape into two distinct futures: 🌑 Legacy BigLaw: - Revenue tied to inputs (hours worked, not outcomes delivered) - Scale driven by associate leverage (junior lawyers billing at senior rates) - Efficiency treated as enemy (faster work = lower revenue) Partnership economics make long-term AI bets impossible 🌕 AI-Native Law Firms: - Fixed, outcome-based pricing at 50% of BigLaw rates - End-to-end automation with 60%+ gross margins - Proprietary datasets that improve with every engagement - Software-like scaling without linear cost increases The math is brutal: A $1.5B firm faces $450M in revenue pressure as AI compresses 30-60% of billable work into minutes. Most vulnerable: M&A diligence, regulatory compliance, patent prosecution, contract lifecycle management. $45B+ in annual fees where "complexity" is often manufactured scarcity. This creates a 10x market expansion - 32M underserved SMBs can now access elite-quality legal work previously exclusive to Fortune 500 companies. The transition is client-driven. GCs are already demanding change: "We expect AI to make things less expensive. Figure it out or we're paying you 20% less next year." ⚡ This transformation represents the largest opportunity in legal services history. ⚡ The next Cravath won't be a partnership - it'll be a platform company with global reach and SaaS-like margins. Let me know if you're building in legal AI. The industry won't have another window this wide open in our lifetime.

1,258

338 Comments

Our recent article entitled "An assessment of synthetic data generation, use and disclosure under Canadian privacy regulations" which just appeared in the journal AI and Ethics evaluates how synthetic data generation (SDG) is treated under Canadian federal private sector privacy law (e.g., issues such as consent and identifiability). Given the increasing adoption of SDG, how the process of generating data and the generated data itself are treated under current statutes and regulations becomes a relevant consideration. However, these statutes and regulations were written before SDG became a prevalent technology. Our analysis can inform future efforts to modernize privacy laws to account for the specifics of contemporary privacy enhancing technologies. The abstract is here: Synthetic data generation (SDG) plays an increasingly important role as a research and innovation accelerator. While SDG can enable privacy-preserving data sharing, it also raises privacy concerns compounded by uncertainty how privacy law applies to SDG and the generated data itself. Such uncertainty can hinder positive applications of SDG and put individual privacy rights at risk. This study aims to understand how SDG and synthetic data are treated under Canadian federal privacy law, identifying regulatory gaps that extend beyond the Canadian context and proposing recommendations to address them. Our analysis shows that SDG is not explicitly addressed by the statute. While SDG arguably qualifies as a use of personal information, it is unclear whether consent is required for SDG. Further Fair Information Practices with respective obligations apply to SDG just as they do to any use of personal information. The generated data itself could fall outside the law’s scope since it is more likely to qualify as non-personal than traditionally de-identified data but the concept of identifiability under the statute remains ambiguous, particularly regarding inferences. An unclear definition of identifiability represents a relevant gap in privacy law that can harm the individual directly, through the exposure of personal information, or indirectly, by hindering the adoption of SDG and other beneficial privacy-enhancing technologies. A Code of Practice, anchored in legislation, could address such privacy concerns and ensure the proper application of SDG.

64

1 Comment

🚦 Levels of Autonomy: A Roadmap for Responsible AI Agents 🤓 Recently we discussed the topic of autonomy of AI Agents in an edition of the Legal Informatics Newsletter (https://lnkd.in/gqt9kt3M). Related tot hat I had several requests for reading recommendations. Here is one on how much autonomy AI Agents should have. 🔖 A recent paper, “Levels of Autonomy for AI Agents” offers insight why autonomy is a design choice, not just a capability. 🔑 The five levels of autonomy: 1️⃣ Operator – user fully in control, agent acts only on request. 2️⃣ Collaborator – agent supports by suggesting or completing subtasks. 3️⃣ Consultant – agent reasons independently but leaves the final say to the user. 4️⃣ Approver – agent executes, but requires human review before final action. 5️⃣ Observer – agent acts fully autonomously; the human just monitors. 📍 In sensitive domains (law, healthcare, compliance), the right autonomy level related to a particular task is a key issue (think of the EU AI Act). ⚠️ Why this matters: 👉 Not every domain can tolerate “observer-level” autonomy. 👉 In sensitive domains calibrated autonomy is essential to safeguard accountability. ✅ The paper introduces the idea of “autonomy certificates” as a governance tool to label and regulate how agents operate, making them auditable and trustworthy. 🎯 Bottom line: This framework provides a scalable roadmap for designing, deploying, and governing AI agents responsibly, ensuring the right level of autonomy matches context, risk, and oversight needs. 🔗 to the paper in the comments #artificialintelligence #law #agent #innovation #future #governance

50

5 Comments

Super excited to share that my latest article is published in the ILTA Peer to Peer Journal (Summer 2025 edition). In this article, I share how law firms can unlock a sustainable competitive advantage by tightly integrating their data and AI strategies. When done right, this creates a powerful flywheel effect, where better data enables smarter AI, and smarter AI generates better customer outcomes and experience, which then creates even more valuable data. This virtuous cycle can fuel continuous innovation, improve decision making, and elevate firm wide performance. The article draws on my experience and learnings at Aderant, where we’re building products and services, such as askMADDI, our conversational AI services, MADDI - our AI OS, and Stridyn, our AI powered NextGen Platform. We’re helping law firms move from reactive operations to proactive, insight-driven leadership. Read the full article here: https://lnkd.in/gTsD68wT Appreciate the opportunity to contribute to the conversation on AI strategy in legal tech, and I welcome thoughts and perspectives from across the industry. #LegalTech #AI #DataStrategy #Aderant #AIstrategy

44

2 Comments

Software vendors love saying: ‘bugs are inevitable’. And they’re right. But here’s the catch, so is accountability! A client once told me: “The software vendor flatly refused to warrant that their product would be defect-free. They said bugs are inevitable.” And they weren’t wrong. Software is messy. Perfection is a myth. But the problem was this: if something broke, the risk shifted entirely onto the client’s business. That’s not acceptable either. So we reframed the conversation. Instead of arguing over “defect-free,” we asked: - What really counts as a defect? (System downtime? Data loss? Security gaps?) - How quickly should each type be fixed? (Critical within hours, minor within weeks.) - What happens if fixes don’t happen? (credits, withholdings, step-in rights, or even termination.) - And how do we ensure accountability isn’t a one-off? (Audit rights, ongoing compliance.) The vendor relaxed once we stopped asking for the impossible. The client was protected because the risk was tied to business impact. That’s the art of tech contracting: not chasing guarantees you’ll never get, but creating levers you can actually use when things go wrong.

41

5 Comments