When Judy Garland skipped down the Yellow Brick Road to try and find solutions for her and her companions’ problems, I don’t recall fixing the house buying and selling process making the shortlist.

Now, I’m not saying a 2023 remake of the Wizard of Oz, featuring a first time buyer of a newbuild in Stockwell is going to give Barbie a run for its money, as it’s quite a niche genre.

I guess the audiences would mostly consist of retired lawyers and surveyors picking holes in the storyline; “No, no, no, that’s not right, Land Registry would never accept an AP1 without the local authority details.” However, it IS a problem that needs to be addressed.

Amongst the completely justifiable barrage of complaints about the time it takes to buy a property, we do come across some who complain that the main culprits are large law firms that are using modern technology, and in particular, artificial intelligence ( AI ), which is only making things worse.

But here’s the thing. Currently, the use of AI in the conveyancing process is almost non-existent, beyond a couple of search firms using rules-based systems for quality control and creating summaries.

The problem is misinformation being spread by those who think there are wizards hiding behind a big green curtain turning wheels and pretending to be, oh, a wizard. But they aren’t.

So what technology are lawyers actually using?

The sad reality is that the technology that most lawyers use is not very smart. The implementation of case management systems is limited, with firms needing to invest the time and money required to customise them to deliver genuine benefits. Many lawyers we hire come from firms where they had Windows folders for documents and if they were lucky, some basic mailmerge facilities.

While there are some areas where the processing has improved, such as onboarding, Stamp Duty Land Tax submissions and registrations, unfortunately, none of these directly impact the time taken to carry out the investigations required. They do save time but most of this work is carried out by assistants rather than lawyers which lessens the impact they have on transaction timescales.

The challenge is that for lawyers, most of the work they need to do is still depressingly manual, and this directly affects transaction times. However, the core difficulty is that this needs to be recognized as a problem that can actually be solved, rather than just be left in the “too-difficult” pile.

A recent comment on LinkedIn typifies how thinking must change. I had suggested that the process of raising enquiries was manual and this could be streamlined. The response was unsurprisingly dismissive; “everyone [ he ] knew was using digital transposing software, which is not manual”.

This translates to a sort of dictation, albeit without the sexist 60’s overtones of “the girls in the typing pool”. The idea that process automation is someone speaking into a microphone which is translated into a pre-contract enquiry is a brilliant example of Oz Wizardry.

It’s not exactly the right solution to the problem, is it?

What will change in the near future and is AI part of it?

Anyone who believes that in the near future conveyancing law firm owners are going to say “Sure, let’s use artificial intelligence to give legal advice and scrap all my lawyers” is being optimistic at best. We need to streamline the process, but the source will have to be rules-based data analysis, and not the use of large language models, currently popularized by ChatGPT.

There will be some incremental changes; possibly how title and search data is returned and delivered to lawyers, more use of electronic data capture and who knows, even some of that material information might somehow wend its way onto the lawyers desktop. The real question here, which people seem to have forgotten, is what are the lawyers actually going to DO with it?

The answer is that they must perform a gap analysis between the issues of concern to the buyer’s lawyer and their lender, and the information that has been provided. The holes in this information would then be highlighted and enquiries raised based on that missing data.

However, to achieve this, we need to firstly convert requirements into rules both for the property, dependent on its profile and information available, and secondly, fully codify lender requirements. The really good news is that the latter is already underway for remortgage work.

However, the harsh reality is that today, anyone who believes that many law firms are using artificial intelligence shares the Scarecrow’s shortcomings and they should probably take a trip down that Yellow Brick Road for some assistance.

Peter Ambrose is the CEO of The Partnership and Legalito – specialists in providing transparent and efficient conveyancing software and services.