Like many industries, law has shifted from dipping its toe in AI to fully adopting it. According to the Solicitors’ Regulation Authority, three quarters of the UK’s largest law firms are using AI, while 30% of smaller firms have got on board.
In a sector that’s known for being traditional and set in its ways, adoption has not been easy. Legal data analytics expert LexisNexis revealed many concerns about its use in modern-day practice. For example, 57% of respondents to a 2023 survey said they were worried about ‘hallucinations’ in AI.
However, the company has also published key insights into the demand for generative AI, drawing on technical experts at leading law firms. Chris Tart-Roberts, Head of Lawtech and Chief Knowledge and Innovation Officer at Macfarlanes, says: “AI is certainly not about replacing lawyers, but we are going to see it turbo-charging their expertise.” And this is our opinion. AI won’t replace lawyers – more likely, lawyers who use AI will replace lawyers.
So where exactly should we expect this to play out?
Practical applications for AI in the legal sector
Generative AI has proven to be gamechanging in an industry that relies on large amounts of data and accuracy. In practical applications, we see it saving lawyers hours of time on:
- Due diligence, such as reviewing contracts and flagging anomalies
- Drafting memos and contracts, identifying risks and suggesting improvements
- Summarising information from invoices to lengthy legal documents
- Completing repetitive tasks in higher volumes
Firms using generative AI to take on these manual tasks will improve their competitive advantage. This is not at the expense of lawyers – instead, legal teams can leverage human expertise for high-value work, such as advising clients and developing case studies.
Risks and concerns about using AI in law firms
First and foremost, the legal sector relies on accuracy. While generative AI is an incredibly useful tool, there have been multiple cases of ‘hallucinations’ in its infancy. An AI hallucination refers to inaccurate or misleading content presented as fact.
These hallucinations can be caused by multiple factors, such as insufficient or biased data to train the AI model. With many law firms slow to adopt due to hallucinations, we need to find reliable datasets.
Poor datasets come with other inherent risks for law, including data privacy and security risks. Many firms simply don’t want the pressure of sharing sensitive data with third parties, causing potential data breaches.
Closed AI models: the solution to legal concerns
Whether it’s a concern about data privacy, security or accuracy, closed generative AI solutions can help to allay them. Forward-thinking law firms are now turning to ‘closed’ or proprietary AI models, which are trained exclusively on legal source material. This means:
- Models are trained solely on reliable legal data, not clouded by exterior sources
- Data is kept private as the models are only available to subscribers
- Lawyers have better control over the accuracy of outputs
- Bias is minimised due to stronger, more refined datasets
- Firms can handle sensitive information safely
While generative AI can improve legal efficiencies, closed AI models can also allay industry-wide concerns. But with manual tasks being outsourced to machines, how have client expectations changed?
How have client expectations changed since the rise of AI?
The rise of AI in the legal sector is no secret. As such, law firms have an ethical obligation to use it fairly, and keep their clients abreast of any changes.
Transparency
To meet these ethical obligations, and minimise risk to credibility, lawyers should be transparent about their use of AI. Many clients have an expectation to be told if AI is used in their case.
Cost reduction
Time savings on manual tasks will doubtless lead to cost savings – and clients are expecting this to be reflected in their bills.
Pricing innovations
This shift in client expectations based on hours saved could lead to pricing innovations. While firms have historically relied on billable hours, they may now have to move to a value-based model.
How can law firms adopt AI sustainably?
These paradigm shifts leave us wondering what’s next for the legal sector, and the AI opportunities within. The most competitive firms will capitalise on the practical benefits of generative AI: saving time processing documents, drafting memos, and reviewing invoices – giving their lawyers better tools to do their jobs better.
But they will also be diligent. In particular, they will draw on ethical sources, using closed models where only accurate data can inform cases and outcomes. This will lead individuals to focus on where they can offer the most value: human-centred advice and legal expertise, not manual processes. Crucially, they will be transparent about where AI has been used.
This will maintain the sector’s traditional focus on data accuracy, without compromising time or trust. In turn, firms will promote efficiency without charging time for time’s sake. AI is not set to replace lawyers, but rather it will empower them, giving clients better value for money and improving their practices.
