Legal academia has tied itself in knots over AI-generated law review submissions, with professors loudly condemning scholars who use AI to draft academic work.
The controversy centres on Pangram, an AI detection platform that flagged a law review submission as containing substantially AI-generated text, triggering widespread outrage across academic social media.
One response to the furore cut to the heart of the matter: “The implication of this moral panic is that law can’t evaluate a quality argument on the merits, without leaning on the sweat equity of keyboard toiling.”
The concern raised is a genuine one, given that legal scholarship has long rewarded scholars for transforming a five-page thought into an eighty-page article of often questionable merit.
Critics have pointed out that Pangram and similar tools are more than capable of delivering career-damaging false positives, raising serious questions about the fairness of using them to screen submissions.
But the deeper problem is that law professors are now deploying another AI tool to identify AI writing they could not otherwise detect, which rather undercuts their stated commitment to authentic human thought.
Professor Panos Ipeirotis of NYU’s Stern School of Business has been running an experiment to illustrate the absurdity, making a short, clear argument and running it through a large language model until Pangram flags it as one hundred percent AI-generated.
He then observes as critics attack the flag rather than engage with the argument itself, summarising the situation bluntly: “It’s a fucking tweet. They are literally using a machine to avoid reading it.”
William Crichton has also weighed in on the broader dynamics of the AI detection market, arguing that “Suspicion isn’t a side effect here, it’s the product being sold.”
That observation identifies the commercial reality driving Pangram’s growth, with the platform now offering a browser extension that scores content in real time for subscribers paying twenty dollars a month or more.
Critics of AI detectors have raised concerns that such tools disproportionately harm writers of colour, with bad-faith actors selectively running minority authors through detection systems they would never apply to white counterparts.
The ableism argument is similarly pointed, with detractors arguing that the product is designed to gatekeep neurodivergent writers who might rely on AI tools to communicate ideas more effectively to a wider audience.
The panic also reflects a deeper anxiety within legal academia about whether editors can still recognise a genuinely good argument when they encounter one, independent of the author’s reputation or institutional connections.
The straightforward alternative proposed by critics is simple: read the submission, assess the reasoning on its merits, and use editorial judgment to determine whether a coherent and valuable argument is present.
If law review editors and professors cannot engage with a written argument without first outsourcing their judgment to an AI detection tool, the problem of intellectual vacancy lies closer to home than they might be comfortable admitting.