Given recent advances in AI large language models which are now used by many law firms as part of their research, clients are understandably asking why this does not translate into lower legal bills.
The answer to this is that, while AI is useful and, some might say, almost essential to assist with tasks such as drafting, research and administrative work, lawyers are professionally obliged to check the results and not leave things entirely to the machines.
This is best illustrated in the recent case of Cork & Anor v Smith (1). In that case, the international law firm involved were rebuked by the judge for their overreliance on AI without proper human oversight.
The case in question was before a judge in the Insolvency and Companies Court. The judge held that certain submissions made to the Court were inaccurate and based upon what he presumed was an AI “hallucination”. He held that, had the submission been checked against the applicable legislation as it should have been, then the error would have been corrected. This was not done and he held that this omission was negligent.
To make matters worse, the judge held that the law firm involved had tried to explain away their error in an untruthful manner which only compounded matters.
Perhaps the most important part of the judge’s decision is this:
“… legal professionals bear ultimate responsibility for their work and cannot outsource the process of legal research or of legal reasoning to an AI. It is a tool to be used with caution. AI has the potential to be wholly unreliable. AI may of course provide a jumping off point for research and legal reasoning but it does not, at least at present, do away with the need for proper research and thought on the part of a legal professional, even a very junior legal professional.”
To answer the question asked by clients therefore – Yes, you will likely see a reduction in legal costs as lawyers become more efficient and the time taken to produce documents and advise on legal issues is much reduced due to the advances in AI which, if anything, make legal research tools more effective and more user friendly. However, AI is not about to eliminate the need for lawyers – or at least not yet. Stepping back, it is worth asking what the client wants from a law firm. Ultimately, the answer is that they want someone they can trust to do what is necessary and right to protect their interests – and crucially someone who has a professional obligation to get it right. AI doesn’t have that. It can’t be sued if it gets things wrong. It can only spew out what someone has put in.
Interestingly the judge’s quote above uses the phrase “at least at present” which suggests that they are ambivalent on whether AI will evolve to such an extent that people will trust the AI more than they trust trained professionals. We shall have to see if that ever happens!
The above decision is a relief for those of us who put in the hard yards to check legal research and make sure things are done properly. However, it’s worth saying that a lot of the cases we handle are subject to arbitration proceedings. In some of those cases, the arbitrations involve foreign law firms who may be uninsured and may not have the same professional obligations as English lawyers. We are seeing increased use of AI in these cases with phantom case references and bald statements being made which are unsupported by legal precedent. It is becoming increasingly difficult to counter some of these arguments as AI seemingly makes it possible to argue the once unarguable. This is a problem not just for us but also for arbitrators who need to devote more time to checking what is said by each side. That all means increased costs – which is the very opposite of what AI is supposed to achieve.
The moral of this story is that AI is becoming essential as a tool to support legal work. However, it cannot replace lawyers completely. It can streamline processes but all outputs must be carefully checked and reviewed by qualified professionals. Furthermore, legal professionals need to be alive to the increasing misuse of AI in legal proceedings and they need to diligently check everything: there is no substitute for qualified expertise.
(1) Cork & Anor v Smith – [2026] EWHC 1199 (Ch)
Source: Court holds that failure to check AI is negligent | BDM Blog | BDM Law