C.BONO.ME

Introduction

Since beginning my legal career in early 2017, I have observed a rapid evolution in legal technology, particularly in artificial intelligence (“AI”). I was lucky to work at two fantastic top tier law firms, Clifford Chance LLP in Australia and Ropes & Gray LLP in the United States.  Both these firms are technologically progressive and are quick to adopt the latest and greatest. Despite strong support from internal technology teams, adoption of legal AI by Associates and Partners was oftentimes dismal – and this is industry-wide.  I hypothesize the following reasons for this were:

• Associates were too busy with billable work and not interested enough to engage with AI.
• The AI tools were cumbersome, inaccurate and largely unhelpful.
• Bad training and use cases.
• Lack of incentives for Associates to use AI (bonuses are tied to billable hours in most cases).
• Partners did not adopt or promote the use of AI.
• Associates felt excluded from the decision to adopt legal AI, while Partners approved its implementation without fully understanding the technology, buying into it or thoroughly testing its utility and capabilities.

Legal AI

In 2022, the release of OpenAI’s ChatGPT marked a dramatic shift in the trajectory of technology.  Unlike earlier AI tools, ChatGPT demonstrated a quantum leap in natural language processing capabilities that captured widespread attention across the legal sector.  Law firms reacted swiftly – some by blocking access to ChatGPT on their internal networks – reflecting both concern over data security and uncertainty about its appropriate use in legal practice.  Many law firms have ethical, insurance, privacy, confidentiality and commercial concerns over the use of ChatGPT in legal practice.

I have a network of friends and colleagues working at top law firms across the world.  Recently, I spoke with friends at top-tier law firms in the US and Australia, and it was striking to hear how little interest their firms have in adopting legal AI tools, testing them, or integrating them into everyday legal practice.  I work at a boutique legal studio, and legal AI dominates our daily discussions because the productivity gains for a boutique firm are so great.

New technology can bring about wholesale change across entire industries.  It is natural for people to be skeptical, especially when they do not understand the technology, or the technology is just completely revolutionary making it very difficult to reconcile with.  Lawyering at the top echelons is technical and multifaceted.  It takes years to be competent, let alone, to master.  AI can be perceived as a direct challenge to the years of dedication and expertise that lawyers have invested in their profession.

At a deeper level, this is existential to legal practice.

Junior Lawyers

AI poses a threat to grueling junior legal work.  Specifically, I am talking about graduates, law clerks and junior lawyers (“Junior Lawyers”), including: (1) blue-collar legal work (process grunt work) (“BCLW”), (2) due diligence (“DD”), (3) simple drafting, and (4) research memorandums.  If completing a buy-side DD report took 100 hours, it could very well now take 20 hours with AI.  This means that lawyers will have to produce 5 DD reports to make up the revenue.  This is fantastic for clients who now get a DD report for $10k instead of $50k or $80k, however, it will no doubt commoditize certain types of BCLW.  Junior Lawyers will no longer have to spend days and nights reviewing and reporting on large contracts in Intralinks VDRs.  There are pros and cons to every technology, but in my view, Junior Lawyers risk developing a weaker understanding of the law and commercial agreements as AI increasingly performs this work for them – a trend akin to “TikTok brain rot.”  Junior Lawyers may be inclined to lean on AI without own independent thought – AI is like brain candy baby!  This is not healthy, but it is very enticing, especially if it is between hanging around the office until 12am every night or using AI and leaving at 7pm instead (arguably, that is healthier!).

This is a debate for another day.

Types of work – AI versus humans

When it comes to front-end legal services – such as mergers and acquisitions, banking and finance and projects – I anticipate these will increasingly fall into three distinct categories, each commanding a different level of value:

• Category 1 – Blue Collar Legal Work (“BCLW”) – Grunt work that is mainly performed by machines and verified, double-checked by junior and mid-level Associates.  Lawyers will need to constantly review and monitor AI outputs to ensure accuracy, relevance and fit for purpose – AI has a competitive advantage at BCLW.  But remember – legal AI does not have professional indemnity insurance!
• Category 2 – High value work – There is another category of legal work performed by Senior Associates and Partners.  This involves greater complexity.  Technical legal drafting, experience-driven decision-making, intangible skills and commercial decisions.  This kind of legal work is difficult to capture with technology due to its technical, multifaceted and non-linear nature.  High value legal work requires complex technical legal thinking – a human competitive advantage.
• Category 3 – Human work – Executive-level legal work that relies heavily on interpersonal skills, professional networks and the ability to navigate dynamic, strategic, opportunity-driven environments; it is inherently human, nuanced and constantly evolving.  A lot of legal work comes down to client preferences, emotional and economic drivers.  Lawyers will need to focus on building strong relationships as a counter defensive to AI, leaning into what is inherently human – another human competitive advantage (bonding, social events, business and industry events).  I am specifically talking about business development, deeply understanding client needs, keeping abreast of industries.

AI will change the way lawyers approach work and clients.  This is not necessarily a bad thing, because it means that lawyers will become versatile and entrepreneurial.  No longer will they be able to sit around for 14 hours per day drafting documents at home and make $300,000 (this is not where legal value will lie in the future), instead, the role of lawyer will change dramatically, and it will become more tactical, technical and executive in nature.

Against this backdrop I wanted to write about the billable hour.  Personally, I do not think the billable hour is a good way to charge for legal services, however, it is a good way to track legal work internally.

As the legal industry moves into an era increasingly influenced by AI, there is growing discussion about the need to reassess or modify the billable hour model to better align with modern legal practice.

Legal AI will reduce the number of billable hours required to produce work.

According to Thomson Reuters, the billable hour began in the early 20^th century, largely credited to Reginald Heber Smith at the Boston Legal Aid Society.  To manage high case volumes, Smith developed a system with Harvard Business School to track lawyer time, improving efficiency and reducing costs.  Over time, this evolved into the dominant model for measuring lawyer performance and firm profitability.  By the 1980s, minimum billable hour targets were common, though the model has since drawn criticism for encouraging inefficiency and misaligned incentives.

Different fee structures

Many types of professional services firms charge by the hour – so it is not just lawyers who should revisit the billable hour.  However, investment bankers (“IBs”) charge fees based on different criteria, such as success on closing.  This is normally a percentage.  Typically, across Australia and the United States, it ranges between 0.5% to 6% depending on the client, consideration or capital raise, deal type, structure and market.  In many cases, clients also pay IBs monthly retainers, as well as other finders’ and referral fees.  This structure is more aligned with value delivered and, in my view, better suited to the future, particularly as AI reshapes professional services.

Below I set out a few different potential fee structures and models:

Final thoughts

I posit that legal AI will upend the business of law, but I wanted to leave you with a few thoughts: