Source: Springbok
The exponential rise of generative AI is causing seismic changes in the legal profession. Princeton University concluded in a survey undertaken this year that legal services are set to be the most drastically impacted by generative AI. Indeed, artificial intelligence poses both the biggest threat, and opportunity, that law firms have ever faced. Those who navigate the risk, make quick decisions, and leverage the technology to their advantage, will ride both the immediate wave of success and come out on top in the long-term.
With a recent European Legal Tech Association survey found that with 9 in 10 lawyers already trying their hand at using generative AI, and 85% supporting incorporating legal tech into law degree curricula, this is clearly a revolution set to stay.
Law firm clients, namely in-house counsel, are actively looking for ways to save costs, access better services, and become more self-sufficient. In fact, 69% of top 100 law firms believe generative AI will have a positive impact on revenues or margins.
The legal industry is known for its conservative approach and prioritising risk mitigation to avoid unanticipated disruptions. Firms are planning to allocate less than £1 million (56%), between £1 million and £5 million (38%), and £6-10 million (6%) into generative AI over the next couple of years. This falls behind consulting firms, which are spending big, such as Accenture’s plan to invest £2.4 million over the next 3 years, PwC to invest £800 million, and Deloitte to invest £1.1 billion.
Currently, the primary goal for most firms is to gain a slight edge (43%) maintain parity (36%) with their peers, or, as with 31% of firms, leverage generative AI to leap ahead of competition in a way that wasn’t previously possible.
As the first step, law firms are designing their preferred combination of collaborating with an AI partner (64%), buying off-the-shelf, and hiring in-house talent.
However, firms generally have zero (44%) or 1-2 (31%) technical members of staff. This significant knowledge and skills gap, paired with a competitive hiring market, is a hurdle to jump over.
No matter the approach, one thing is clear: law firms’ fear of becoming irrelevant is high, and visionary executive teams are concluding that the risk of not taking action far outweighs the risk of taking action. Moreover, first movers will positively impact their client portfolio and attract new clients in AI.
Very few firms (6%) are solely relying on in-house talent recruitment for their execution strategy. However, the biggest constraint (64%) impeding firms from building out their own solutions is lack of capacity and of strong technical delivery talent.
In fact, the work, and team, involved in taking a data science PoC experiment into production is very frequently underestimated, and by a big margin. This explains why so many law firms appear to be working on AI projects, but have yet little execution or firm-wide roll-outs to point to.
The execution gap can be explained by two factors: Firstly, law firms focus on what they are uniquely good at: the law. As such, AI delivery teams within law firms are almost non-existent. Secondly, law firms have historically struggled to attract top tech talent – driven by less competitive salaries than Big Tech, lower calibre engineering mentorship opportunities, and limited upwards career growth.
Due to the talent shortage, firms are reportedly ‘poaching’ entire teams of data scientists from rival firms.
While many firms do have some technical talent in-house, these are often constrained by maintaining internal legacy solutions, firm politics, and lack of resources to move data science experiments into enterprise-grade production.
The most widely-piloted off-the-shelf generative AI solutions used by lawyers to date are Microsoft 365 Co-pilot (50%), and Harvey (31%), CoCounsel (31%) and fleetAI (13%).
Despite the popularity of pilots as a starting point, so far, few firms have onboarded Generative AI solutions. Only a small fraction – approximately 5% of law firms – have opted to exclusively purchase off-the-shelf solutions without internal or external technical personnel. Given lawyers’ trepidation around inputting client-sensitive data into a solution hosted off-site, they are seeking specialised products that use their unique data but also with mitigated data privacy risk.
In the process of building a generative AI product, the initial planning steps are technical feasibility assessment, detailed solution scoping, user journey mapping, and a development roadmap. The next step is an agile methodology driven workflow consisting of product management, stakeholder alignment, product development, LLM operations and development operations, and model development and improvement. During the subsequent period, the pilot must be managed with continuous adjustments, in the lead-up to deployment and the roadmap expansion that follows.
Generative AI-based solutions are enabling profitability in alternative fee arrangements, or AFAs. Fixed fees are by far the most common alternative fee type (offered by 96% of firms), followed by stage-based fees. Law firms have used AFAs for over 10 years already, with 85% of firms reportedly using AFA-based flat fees because clients continue to demand it, and increasingly so, with 4 in 10 lawyers stating they are seeing a surge in client demand for AFAs.
Making capped fee arrangements consistently profitable has been a longstanding challenge, but a survey by AltFee found that AFAs now are more profitable than hourly pricing. This is contrary to the previous general consensus that AFAs are not profitable for law firms, and are only adopted out of pressure from clients. The reality is that poorly executed AFAs are not profitable.
Lawyers predominantly process their work manually. To date, the only arbitrage available to law firms was in leveraging talented junior staff for high volume execution at lower cost. However, it is common for legal practices to write off up to 30% of their total hours on matters, and to achieve just 3-10% annual profitability. Until now, without leveraging generative AI technology, this margin could only be narrowly improved.
Increasing margins is all the more reason law firms are now looking at AI partnerships. However, if law firms are not sufficiently reactive, and resist the movement away from billable hours, then both their profitability and client portfolio could be negatively impacted.
A survey conducted by Wolters Kluwer in 2022 saw that more corporate legal departments than the previous year are considering switching law firms, and a significant reason cited is efficiency and productivity. Over 90% of corporations’ legal departments consider it important for law firms to fully leverage technology, and 70% are reportedly asking law firms they’re considering working with to describe the technology they’re using.
In 1-2 years’ time, the forward-thinking law firms of today will be releasing their domain-specific, prompt architected solutions to the market, fundamentally altering pricing models by incorporating technology subscriptions. Within 3-5 years, the traditional billable hour in corporate law will become obsolete for routine non-contentious, non-bespoke work, as forward-thinking clients demand alternative approaches.
Generative AI offers opportunities to imagine new ways of delivering legal services in a manner unimaginable a few years ago. As such, law firms are faced with a double-edged sword of risk and opportunity. Firms that can deftly navigate this evolving terrain will not only seize immediate success but also shape the future of legal practice.
