Is AI a threat to the work of lawyers and judges?
I talked about this with Irina Carnat.
Irina is a postdoctoral researcher at the Lider-Lab and l’EMbeDS of Sant’Anna School of Advanced Studies and author of the well-written paper ‘Automation as Delegation of Power: Constitutional Constraints on AI Systems for the Administration of Justice’.
The narrative around AI replacing lawyers is flawed. Lawyers obtain value from AI systems through their own interpretation and understanding of the system’s output. LLMs don’t do anything on their own. Therefore, AI is not a competitor to lawyers, but a collaborator.
Even though law firms are probably required to adopt AI to stay competitive in the future, this will change the legal profession slowly over time, not rapidly disrupt it overnight. Regulatory safeguards ensure that meaningful human control is in place to avoid “cognitive surrender” in task delegation to AI.
Yes, AI does pose a threat to the future job prospects of junior lawyers, but maybe fewer, but better lawyers is not that bad after all.
We should also address the common claim I see on LinkedIn and Substack that EU’s regulation presents a hindrance to innovation. That is only the case, if “innovation” means that companies can spend less and increase profits. From an EU perspective, if “innovation” does not support fundamental rights protection, that is not the kind of innovation we are looking for. Instead of loosening regulatory safeguards, we should invest more in better legal scholarship to help companies transition through the compliance process.
See the full conversation here on Substack, via my YouTube channel, or listen to it on your favorite streaming service, and let me know what you think in the comments below.
Links:
AI will kill all the lawyers (article in The Spectator from December, 2025)
Moral Crumple Zones: Cautionary Tales in Human-Robot Interaction
