SMH: ‘AI won’t kill the law degree. It will redefine it’

This Sydney Morning Herald op-ed reads as marketing copy dressed as analysis. Two law school administrators defend law school. That is the structure and the limit.
My four coalition questions apply. Johns and Walton draw their salaries from Sydney Law School. Their authority rests on the premise that law schools produce irreplaceable skills. Prospective students pay the tuition that funds their chairs. A piece arguing law school might not be worth it cannot be written from their positions. The byline is the argument.
The argumentative move is Turner territory. When a profession feels its jurisdiction threatened, its defenders claim the work rests on tacit knowledge that resists formalization. Johns and Walton make the move directly. AI “cannot weigh authority and precedent with deep contextual understanding,” cannot “navigate jurisdictional boundaries,” cannot “appreciate distinct legal subcultures.” These claims are asserted, not shown. The evidence runs the other way. Document review, legal research, drafting, case analysis, contract markup — the bulk of junior lawyer work — are exactly what current models handle. Every limit the authors cite in April 2026 existed in April 2024, and several have eroded in the interval. The piece treats AI as a fixed object rather than a moving front.
The authors never confront the pipeline problem. Even if senior lawyers retain judgment AI cannot replicate, the training ladder runs through paid junior work. Strip out that work and graduates have nowhere to stand. The article says nothing about this because the authors cannot say anything about this without threatening the product they sell.
The “five things” section is a curriculum prospectus. It tells prospective students that Sydney Law School is adapting, that its graduates will be techno-legally fluent, that the degree will pay. That is what a dean says. It is not analysis.
The revealing sentence is “We believe it is.” Of course they do. Their mortgages depend on believing it.
A more candid piece might address starting salaries relative to debt, automation of document review and its effect on training pipelines, whether a philosophy or history degree cultivates the same “human capacities that resist automation” at a fraction of the cost, and what happens to Australian law graduates over the next decade if Suleyman is half right. None of that appears.
The one patch of something closer to thought is the gesture at political economy. They note that AI development depends on capital, energy, minerals, tax incentives, and that the corporate structures of OpenAI and Anthropic shape who holds technological power. That is the closest the piece gets to honesty. But they fold it back into the sales pitch: future lawyers will master these questions, and Sydney Law School will teach them. The possibility that the credentialing apparatus itself serves the interests its graduates assume it serves never surfaces.
The prose also does the thing their own argument warns against. They ask graduates to evaluate AI outputs “discerningly rather than defer to them.” The article asks readers to defer to the authors’ discernment about the value of the thing the authors sell.
A Becker read would call law school a hero system under threat, and this article a ritual reaffirmation for members of the cult. The ritual works on the faithful. It will not persuade a skeptical eighteen-year-old sitting with an ATAR and a calculator.