AI promises faster, better and cheaper legal work. But are lawyers placing too much faith in its output and neglecting their responsibility to verify it?
We need to start moving away from the hallucinated cases discourse sooner rather than later.
I am a bit perplexed by how much attention lawyers continue to give to hallucinated cases.
Before AI, it was always our responsibility as lawyers to check the cases and citations we relied upon. It was part of the process. That doesn't mean it was always done properly, but the process and responsibility were clearly there.
So why does it seem to be such a problem now?
Is it the promise of faster, better and cheaper legal services, which AI purports to deliver and consumers and clients have embraced, combined with our blind faith and overconfidence in what AI spits out, that is leading us to trust the output more than we should?
Perhaps we want to believe so much in what AI can deliver that we are overlooking what we still need to do.
AI can produce entirely fictitious authorities, quickly and convincingly. That is a genuine risk and one we cannot ignore.
However, that now makes verification more important than ever before, not less.
If we are sacrificing verification in response to pressure from consumers and clients for faster, better and cheaper legal services, are we really delivering any of those things?
The problem is not just AI.
It is how we are using it and what we are expecting it to deliver.
It is also a communication issue with consumers and clients.
We need to be clear with them that AI-generated work still requires proper checking and verification, potentially at a level we have not previously been accustomed to.
Lawyers have always been the checkers and gatekeepers of the legal work that goes out the door.
AI hasn't changed who is responsible for checking it.
