Reading time: five minutes

Despite the efforts to move the legal business model towards alternatives, the billable hour remains the predominant model of law firms globally.

However, with the arrival of generative AI, the legal business model is beginning to be challenged.

What’s generative AI?

Generative AII is artificial intelligence capable of generating text, images, videos and other data using generative models usually in response to prompts. It learns the patterns and structure of its input training data to generate new data that has similar characteristics.

The most well-known generative AI systems in recent years are the ChatGPT chatbot and image generation systems, such as Midjourney and DALL-E. The user feeds a prompt into the system, such as a question or an image description, then the system generates a response based on the data it has already been programmed with to give the impression of intelligence.

Before considering the implications for legal business models, let’s first have a look at what we have already.

The billable hour

Think of the billable hour model like the LPC next to, say, the fixed fee model, which you could liken to the SQE. Basically, the billable hour is tried and tested because of its simplicity and familiarity, just like the LPC has been the familiar and tested way of preparing aspiring lawyers for the practical side of working as a lawyer. Whereas the fixed fee model is still fairly new, like the SQE, and still needs time to get established as the norm going forward.

However, when you step back and analyse this further, the billable hour system isn’t a good measure of worth or productivity because it can give the impression that legal professionals can take their time on simple tasks to internally measure their worth to their employer, as well as more easily hit their billing targets. Just because a lawyer is hitting their billing targets using this model, it doesn’t mean that it translates into a good and timely service for their clients.

Time recording

In theory, tracking your time by the six minute unit allows you to make an accurate invoice for the client at each agreed period or the end of the matter. However, even products that claim to track time accurately don’t take into account things that are taking place off the device they are linked to, such as taking time to answer questions from trainees or other staff, bathroom breaks or food and drink breaks.

With this in mind, what are the implications of generative AI for law firms and the legal business model?

AI taking on more routine tasks

The most common implication is generative AI taking on tasks that would normally be allocated to junior lawyers, such as:

• handling initial enquiries;
• document review;
• research; and
• drafting.

A generative AI system being able to do this quickly and accurately will streamline operations and improve efficiency, freeing up time and capacity for lawyers to focus on the more technical work that an AI system can’t do.

Another upside of this is that junior lawyers, particularly trainees and NQs, will be able to receive more technical training on aspects of legal service provision that AI systems can’t do, including:

• risk management;
• business planning;
• client pitches; and
• business development.

Alternative fee arrangements (AFAs)

Because of the opportunities that using generative AI can bring to law firms and their clients, firms can start to consider AFAs as part of a move towards greater cost predictability and transparency for clients.

Some law firms have already adopted this model as their primary billing model, including some that I’ve previously worked at. The most common ones I’ve seen in practice are the fixed fee model and the subscription model.

Fixed fee models are where the client is given the costs from the outset based on what they need from their lawyers. For instance, if a client needs a sale and purchase agreement, the firm will give them a fixed price based on the value of the service it’ll be providing. The firm may also give the hourly rates of the people working on the matter if it turns out that further work outside of the scope of the initial fee quote is needed, but it isn’t the primary model here.

Subscription models are where a client pays a periodic subscription, either monthly or quarterly or another frequency agreed with the firm, and the client has access to the firm’s services. The agreement between the client and the firm will go into more detail about the quantity, complexity and type of service they have access to – such as document review, business planning or even unlimited legal advice.

But in an AI-driven legal world, will we still need lawyers?

Of course we will! AI won’t replace human workers, despite what many influential people might tell you. With AI taking on routine tasks, lawyers’ time would be free to invest into higher-value work and training, they can complete matters more efficiently, which will result in a better deal for the client, and we can invest more into deeper lawyer-client relationships.

Going forward, try out periods and training staff on how to use AI systems will be a good start for a lot of firms. Doing this in a controlled environment means disruption to normal workflow can be minimised and, if the try out period is a success, the firm can look at implementing the AI systems across the rest of the company.

The gold standard that lawyers should be aiming for is to be trusted advisers. So, it makes sense to bring in measures or services that can help us do just that so that we then have a much greater capacity to do the job to the highest possible standard.