4 Questions – how would Professor Richard Susskind answer?
1. How will emerging Legal AI solutions reshape the delivery of legal expertise?
Susskind’s answer would not be: “lawyers will do the same work faster.”
His consistent position is that the deeper change is not automation but redesign. Legal AI will first improve productivity, but its more important effect will be the creation of new legal delivery systems: guided self-help, legal triage, AI-supported contracting, online dispute resolution, dispute containment and eventually dispute avoidance. In his own framing, the long-term impact is not replacing lawyers task-by-task, but delivering legal outcomes in new ways.
He would likely distinguish three phases: short-term AI as productivity support; medium-term AI as a client-facing legal guidance layer; long-term AI as infrastructure for prevention, resolution and enforcement of legal rights. This fits his repeated argument that current commentary overstates near-term disruption but understates long-term transformation.
2. What will clients expect as Legal AI moves into practice?
Susskind’s core answer: clients do not primarily want lawyers; they want legal outcomes. In one of his recurring analogies, “people don’t want doctors, they want health.” Applied to law, clients want risk reduction, enforceable rights, avoided disputes, faster resolution, lower cost, better transparency and business enablement.
For corporate clients, the expectation will move from bespoke lawyer-led service to fast, embedded, auditable legal capability inside business workflows. The Amsterdam University of Applied Sciences report of Susskind’s 2025 remarks captures the point sharply: AI-driven tools can become competitors to law firms because citizens and businesses may handle many legal matters independently, from contracts to dispute resolution.
3. How will the role of lawyers and legal institutions evolve?
Susskind would not say that all lawyers disappear. He would say the traditional lawyer ceases to be the unavoidable intermediary. In the short term, lawyers are “turbocharged”; in the longer term, non-lawyers, businesses and citizens gain access to systems that do much of what lawyers historically mediated.
The lawyer’s role shifts toward system design, legal knowledge engineering, risk architecture, validation, ethics, governance, strategy and complex human judgment. He also expects new legal roles, including legal data scientists and legal knowledge engineers. Legal institutions, especially courts, must move from building-centred, lawyer-intelligible procedures toward accessible digital front ends, online courts and redesigned public justice services.
4. What should legal leaders do now?
His practical answer is: do not merely buy tools; rethink the service model. Leaders should identify the outcomes clients and citizens actually need, then ask which parts can be automated, which can be innovated, and which can be eliminated by prevention. Susskind has warned against “technological myopia”: judging future AI by today’s defects. He also warns against “not-us thinking,” the tendency of each profession to assume AI will transform every field except its own.
For law firms and legal departments, the immediate agenda is: experiment seriously, build trusted data and knowledge assets, redesign workflows around clients, train lawyers for AI-native practice, create governance for accuracy/confidentiality/accountability, and prepare for competition from systems that deliver legal outcomes without traditional lawyer involvement. For legal education, he is especially critical: he argues that many law schools still train lawyers for a late-20th-century model of practice.
Strategic synthesis
Susskind’s message will be: Legal AI is not mainly a tool for making lawyers more efficient. It is a catalyst for rethinking how law is accessed, delivered, priced and trusted. The winners will not be those who simply add AI to existing legal processes, but those who build new legal systems around client outcomes, prevention, accessibility and verifiable trust.
Strongest counterargument
The strongest counterargument is that current Legal AI remains unreliable for many legally significant tasks, especially where facts are incomplete, legal sources conflict, discretion matters or the answer cannot be easily verified. Recent research cautions that the most transformative claims about legal AI are also the hardest to evaluate and most exposed to over-optimism. That does not refute Susskind’s long-term thesis, but it supports a phased, audited, human-supervised deployment model rather than uncontrolled substitution.