Series introduction

Artificial intelligence is not simply another technology being adopted by the legal profession. It is structurally reconfiguring the profession — how legal services are delivered, who delivers them, how lawyers are trained, and what it means to exercise professional judgment. Understanding this transformation requires more than tracking which firms have adopted which tools. It requires seeing the legal profession as a system — one in which multiple forces interact in ways that no single vantage point can capture.

We come at this from different directions, and that's the point.

Anthea is an international law and global governance professor at ANU and a visiting professor at Harvard Law School. She also builds and orchestrates AI tools for complex decision-making — not in the legal space specifically, but across risk, strategy and policy more broadly. She is interested in how agentic AI can augment human decision-making in high-end tasks, rather than simply automating routine ones. She follows AI developments closely and reads market moves like Harvey's $11 billion valuation as technology strategy plays.

David is the Lester Kissel Professor of Law at Harvard, Vice Dean for Global Initiatives on the Legal Profession, and Faculty Director of the Center on the Legal Profession. He has studied the legal profession as a system for decades — who enters it, how careers unfold, how firms make money, how the profession globalizes and transforms. He reads those same moves in AI and legal tech as the latest chapter in the century-long evolution of how legal services are organized and delivered.

One reads the tech strategy. The other reads the professional structure. Together we see things neither sees alone.

This project rests on three foundations. Each represents a stream of work that one or both of us have been pursuing independently. Together, they frame why we think this project matters and what it's trying to do.

Complex Decision-Making in a VUCA World

The first foundation is a shared conviction — arrived at from different directions — that navigating complexity requires integrating multiple perspectives, not retreating into narrow expertise.

David has observed this in the legal profession over decades. The best lawyers — and particularly general counsel — don't just do legal tasks. They operate at the intersection of law, business, risk, strategy, politics, and public relations. The partner who closes a transformative merger isn't doing "law" in the narrow sense. She's integrating financial analysis, regulatory risk, employment implications, and reputational dynamics into a single coherent recommendation. The general counsel who steers a company through a geopolitical crisis isn't applying legal doctrine. She's exercising judgment across domains that no single specialism covers. The top lawyers have always been paid for this strategic, integrative outlook. But as the world has become more volatile, uncertain, complex, and ambiguous — what strategists call a VUCA world — this integrative capacity has moved from a nice-to-have to the core of what the profession needs.

Anthea noticed the same thing in debates about international law and globalization. But she took it one step further by building out a methodology around exactly this kind of integrative thinking. What Anthea calls "dragonfly thinking" draws on Philip Tetlock's research on expert forecasting, which found that the best forecasters aren't deep specialists but integrative thinkers who see through multiple lenses simultaneously. The dragonfly metaphor captures it: a dragonfly's compound eyes integrate thousands of lenses into nearly 360-degree vision, enabling it to understand its environment and adapt with extraordinary agility. Dragonfly thinking, as a methodology, involves synthesizing a multitude of points, counterpoints, and counter-counterpoints so that the integrated whole is more valuable than the sum of its parts.

The convergence between these two streams is what brought us together. David saw it in the profession — the best lawyers are dragonfly thinkers, whether they use that language or not. Anthea saw it in global governance and policy-making and built a methodology around it — dragonfly thinking as a systematic approach to navigating complexity by integrating diverse perspectives into compound vision. The observation and the methodology went hand-in-glove.

AI for Complex Decision-Making

The second foundation is about what happens when you bring AI into this picture — not as something that threatens complex decision-making but as something that might make it better.

Anthea's work at Dragonfly Thinking (a start-up she founded to build her AI method) has produced a methodology applied through AI agents designed to help decision-makers understand and act in uncertainty. These aren't legal AI tools. They're a method for AI-enabled multi-lens analysis across risk, strategy and policy — systems that help leaders map actors, trace drivers, run scenarios, challenge their own assumptions, and surface blind spots they didn't know they had. The tools operationalize the dragonfly thinking methodology: they help humans hold more perspectives and system dynamics in mind than unaided cognition allows. They are also designed to enable interdisciplinary teams to see each other's perspectives and to align on what to do.

Together, we've brought these tools and frameworks to senior decision-makers in the legal profession. In 2024, we co-led a senior leadership workshop at Harvard with chief legal officers from Blackstone, Vanguard, BlackRock, and AIG, and managing partners from Morgan Lewis, Orrick, and McGuireWoods. We talked through the frameworks underpinning this work — such as the Risks, Rewards, and Resilience framework for mapping complex decisions, the dragonfly methodology for integrating multiple lenses — and gave demonstrations that allowed participants to interact with the AI tools to think through complex decision-making problems. We've continued through speaker series on dragonfly thinking for in-house lawyers and through the ongoing conversations with practitioners that inform everything we write (see also David's interview with Harvard Law Today on AI, uncertainty, and the demand for lawyers).

This gives us a distinctive vantage point on the subject of this series. We are using AI tools to analyze AI's impact on the profession and we will be sharing a series of papers and visualizations we have co-created with various Dragonfly agents. We're not writing about the AI experience from the outside. We're living it from the inside — experiencing firsthand the gap between model capability and domain expertise, the moments when AI surfaces something genuinely surprising and the moments when it produces fluent nonsense that only an expert would catch. That practitioner's understanding feeds directly into our analysis. When we write about the ingredients needed for good AI outcomes, we're drawing on our own experience of trying to get them.

How AI Restructures the Legal Field

The third foundation is what emerges when you bring the first two together. What happens when AI — with all its potential to automate routine work and enhance complex decision-making — meets the structural position that law has built over a century?

We both see the personal anxieties this provokes. We see it in law students asking whether their degrees will be worthless by the time they graduate, and in senior lawyers wondering whether decades of hard-won expertise are about to be devalued. Anthea sees it when teaching people to use AI tools — some see it as an opportunity and pick it up quickly, while others view it as a threat and find it hard to adapt. These reactions aren't irrational. They're responses to genuine uncertainty about what expertise means when machines can do things that used to require years of training.

But those individual anxieties don't stay individual. They ripple outward — from the person to the firm to the field. A partner who resists learning AI shapes a practice group's culture. A firm that delays adoption shifts its competitive position. A profession that underinvests in training its next generation risks hollowing out the expertise on which its authority rests. We like to trace these dynamics across scales because AI is not just affecting individual lawyers — it is reshaping the system of the legal profession at every level. And multi-scale analysis and system dynamics are core features of dragonfly thinking.

David's decades of studying the legal profession as a system also reveal that law occupies a more central, more connected position in how business and society operate than most people — including most lawyers — recognize. AI is now arriving in a profession with that structural position. The implications go well beyond what it will be like to practice law or how individual lawyers' roles will change — though both matter enormously. They extend to the business model of law firms, which players are in the market, the future of the billable hour, and whether law's structural centrality holds, expands, or gets captured by the platforms being built on top of it.

This third foundation is what the essay series explores. Not complex decision-making in isolation. Not AI in isolation. Not the legal profession in isolation. But what happens at their intersection — and what that means for lawyers, firms, clients, and the profession's future.

The Series

The pieces in this series are ideas and frameworks we're developing as we co-teach a course on AI and the Future of the Legal Profession at Harvard. We drafted the opening essays together in Cambridge at the end of March, while teaching the course and hosting a colloquium on agentic AI and complex decision-making. The series continued in early April from opposite coasts — Anthea in Silicon Valley, David in Cambridge — as we processed and extended what we'd explored together. We share the pieces as they emerge — not as finished pronouncements but as lenses being tested and refined.

The first three pieces in this release each take a different cut at the landscape:

1. Law is the gateway drug argues that law's structural position — at the center of business, connecting every other function — makes it the natural entry point for AI across all professional services. The question: does law capture the gateway position, or does the gateway capture law?

1. Harvey's Strategic Evolution examines what the most-discussed legal AI company actually appears to be building — and why reading it as a "ChatGPT wrapper" misses the strategic thesis entirely.

1. Getting your hands dirty turns inward — from analyzing the system to experiencing it. What does it actually feel like to work with AI agents on complex intellectual tasks? What blocks the profession from getting there? And why does the gap between talking about AI and working with AI matter more than most people appreciate?

These pieces connect. The gateway drug essay establishes law's structural centrality — why it touches everything and why capital is flooding in. Harvey's Strategic Evolution shows a company trying to build on that structural position. Getting your hands dirty turns inward to ask the most immediate question: are you willing to learn the new medium?

Three further pieces will appear in the next release: The Scrambled Competitive Map (the competitive landscape AI is redrawing — incumbents, challengers, and clients all exerting gravitational pull); The Training Crisis (the four ingredients of good AI outcomes, and why the profession is systematically undermining the scarcest one); and From Up or Out to Up and Out (where elite lawyering goes when AI automates the base of the pyramid — and whether the destination is reachable when the ladder behind it is collapsing).

What Comes Next

This project is not just a paper series. It is a sustained program of research and engagement at the intersection of the three questions above — questions that are going to affect every part of the legal profession, law firms, and legal education for years to come.

In addition to the essays and interactive visualizations shared here, we are developing a course at Harvard Law School on AI and the future of the legal profession, and we plan to reconvene workshops with senior leaders in law — general counsel, managing partners, and others at the frontier of these changes — to continue testing these frameworks against the experience of people living through them. In April 2026, we hosted a colloquium at Harvard on agentic AI and complex decision-making, bringing together researchers and practitioners to examine how agentic AI could enrich research projects.

The topic network visualization maps the corpus of sources we are building as part of this project — at the time of this release, 170 sources across 39 topics in 11 dimensions, from market dynamics and technology evolution to competitive dynamics, human-AI collaboration, and professional identity. Each source is coded against these dimensions, creating a map of which issues are most discussed, where structural tensions live, and which sources bridge multiple domains. This is an ongoing, living collection — the knowledge base that underpins the analysis in this series, refreshed through monthly sweeps as new developments emerge.

Future papers will take the analysis deeper — mapping the actors reshaping the legal market, the driving forces behind their strategies, the feedback loops that amplify or constrain change, and the scenarios that emerge from different combinations of these forces. The papers and visuals are shared as they are developed — not as finished pronouncements but as frameworks being tested against the evidence and refined through engagement with the profession.

We're interested less in predicting who wins than in helping people see the system clearly enough to make better decisions within it. We also welcome input from the broader ecosystem we're addressing — if you have thoughts on these pieces, if you disagree with some of the analysis, if you want to say "yes, but ..." then let us know. This is an evolving project and it takes many lenses to create dragonfly vision.