Judgment writing in the age of AI

I’m a big fan of Justice Kós’s writing. Clear, vivid, and accessible. One of my favourite examples is a drink-driving appeal, which starts:

“A bad day for Mr Daniels, and one about to get worse. After drowning his sorrows at a local bar he went out to his truck and started the engine. At that point a police constable emerged from the shadows of the night, breath-tested Mr Daniels and arrested him.”

Which is what makes the Supreme Court’s recent decision in Kea Investments Ltd v Wikeley so interesting. For the substance of the decision, which turns on the interplay between fraud and comity, have a look at my earlier post (link in comments).

But this post is about language. Justice Kós wrote the Court’s judgment. Unsurprisingly, it is clear and well written. But he explained the key issue with the phrase “fraud necessarily reframes ordinary comity considerations”, variations of which he used four times.

A couple of years ago, I wouldn’t have given this a second thought. In fact, I probably would have seen it as a nice turn of phrase to explain the Court’s reasoning.

But we live in an AI world, and most readers are by now well accustomed to AI churning out answers that “reframe” a point after “delving” into a “landscape” that is “intrinsically multifaceted”.

This is not a comment about the use of AI in the courts. Justice Kós wrote this judgment, not AI. It is an observation about how AI has worn out so many words and phrases that I now review my own writing and ask: does that sound like something AI would say? Because if it does, out it goes -- even though I wrote it myself.

I’d be interested to know whether other people are doing the same.