Goodbye job applications, hello dream career
Seize control of your career and design the future you deserve with LW career

Are BigLaw firms using AI to its full potential?

As emerging technologies and generative artificial intelligence (AI) become commonplace in legal workplaces, firms can now gain a competitive edge by using new tech to reduce costs and provide innovative solutions to complex legal issues. Here, BigLaw partners discuss whether the potential of AI in firms is being fully utilised – and if, moving forward, AI should be a key part of a law firm’s strategy.

user iconLauren Croft 24 April 2024 Big Law
expand image

Over the course of last year, AI tech and platforms like ChatGPT made global headlines and prompted various debates across the Australian legal profession.

You can read Lawyers Weekly’s full coverage of ChatGPT and other AI platforms and what lawyers need to know here.

While there is still some fear in the profession that the rise of such tech could mean the beginning of the end of lawyers, clients are demanding efficiency more than ever. However, the legal profession has reportedly not embraced emerging tech at the same pace as other sectors.


However, despite being called “no different to Wikipedia” by some, AI tools can be a “useful resource” for boutique and BigLaw firms alike and will require a focus on key skills moving forward.

Far-reaching uses and benefits of AI

After the rise of ChatGPT, many BigLaw players began implementing similar AI tech into their practices: KPMG partnered with Microsoft to develop “KymChat”, a proprietary version of ChatGPT, and both Allen & Overy and PwC partnered with AI platform “Harvey”.

And in February last year – in a first for Australian law firms – Clayton Utz built a product to help its lawyers use and benefit from ChatGPT in a “reliable and effective” way. The product initially works with another commercial AI tool to pull key information from a case. Once complete, ChatGPT converts that output into a case summary. This summary is then reviewed by a lawyer for any required edits.

Clayton Utz forensic technology and data analytics partner William Howe said that for the firm, having a “client-centric” view towards AI has been helpful when deciding what initiatives and tech to pursue.

“Some Australian firms, including ours, have been transforming over the past few years with new business models that incorporate technology right into the heart of the practice. We have seen reward on that investment and have been able to move quickly around AI,” he said.

“We will be the first law firm in Australia to access Lexis+AI, and we are trialling Microsoft Co-Pilot [and are] developing a comprehensive suite of client-first applications that use generative AI alongside our lawyers to solve complex legal challenges.”

For BigLaw firm Gilbert + Tobin, AI has assisted greatly with document reviews, finding and extracting essential data, and by providing rapid insights into large data sets. Firm partner and chief knowledge and innovation officer Caryn Sandler said the potential of this tech is yet to be realised.

“We are currently actively, and carefully, experimenting with new and emerging AI technologies, including generative AI, to identify ways to achieve greater efficiencies for our staff and clients across all practice areas. In our experience, the use of AI technologies historically in areas of e-discovery and expert systems has been very successful. However, we have found machine learning to be underwhelming in areas such as in contract review, meaning there is undoubtedly potential for GenAI to provide uplift in areas like these,” she explained.

“What is clear to us is that the potential use cases for AI will continue to expand, and as a consequence, all law firms need to continue, and adapt, their AI experimentation, testing, and use cases. The more we test the limits of these models, and learn of improvements to them, the more we expand our thinking on how we can use AI and what its potential for us, and law firms more broadly, might truly be.”

Piper Alderman chief operating officer Chris McLean also confirmed the firm has been trialling AI-based systems in several areas, including developing its own in-house models using Azure GPT – a Microsoft multimodal model.

“Although many firms are making big claims regarding the benefits of AI in legal practice, the reality at the moment is that the major areas of transformation are in business development, marketing and administration,” he explained.

“The next generation of AI products that are only just emerging are likely to change this, with more general legal AI tools from the big publishers and specialist AI tools targeted at M&A, large litigation and other practice areas. At the moment, AI has a strong value proposition in low-margin, high-volume work like workers’ compensation, where most of the processes can already be automated, and AI can bring this to another level.”

However, while GenAI and emerging tech have been a key legal trend and “hot topic” of discussion lately, Mills Oakley partner Dalvin Chien noted that “analytical AI tools have been in circulation and utilised by many law firms and legal departments alike long before this”.

“Often, I solve problems where there is no clear legislation or clear regulatory guidance. We are also often asked to apply a level of ‘crystal-balling’ to our advice to account for emerging technology, including issues with artificial intelligence itself,” he said.

“In its current form, many of the generative AI tools are limited in providing me [with] the answers I need. There is, however, a use case for the use of generative AI for base-level research. There is also a strong use case, I think, for document analysis and preparation, especially with current iterations of the tool which addresses key issues associated with, for example, data security.”

Challenges and opportunities

Despite a recent report from Dye & Durham revealing that in 2024, firms will turn to AI more and more in a bid to keep talent and drive efficiency, there remain a number of key challenges around the implementation of emerging tech into law firms.

“What currently seems to be out of reach for AI are big, complex, novel legal issues. These require understanding of context, human dynamics, broader society, and skills like empathy and negotiation. The best computer we have for these types of problems today is a lawyer. Who knows what the future will bring, but for now, the humans are firmly in the lead,” Howe added.

“Although the past year has brought tremendous change to the world of AI, the first iterations of solutions in this space solve the same types of problems that previous AI always has: large-scale repetitive tasks. The workflows that currently seem to benefit the most are those which have some element of scale, which allows for appropriate time to be spent in refining the approach and for clients to see a return on that investment.”

However, the full extent of these challenges is still unknown, as GenAI technologies “are still in their infancy”, according to Sandler.

“Despite the rapid improvements in these technologies, there are well-founded concerns within law firms around the accuracy of the output of GenAI chatbots. It is now widely known that these technologies are prone to ‘hallucinations’ (where the AI generates false or nonsensical information). While the latest GenAI models seem to be doing better at reducing the number of hallucinations, their higher-quality output means that it can be harder to spot those hallucinations. This means that firms must continually check, and recheck, the accuracy of GenAI outputs,” she added.

“As with all technologies firms deploy when delivering legal services, firms are responsible for equipping lawyers with the skills, tools, and presence of mind to remain vigilant and to ensure that they are not delegating their judgement to the technology. In addition, law firms must always consider broader ethical obligations, data protection issues, and client requirements.”

Despite this, “there are significant opportunities for operational efficiencies” for firms utilising AI, McLean said.

“The main challenges with AI are the fundamental issues with current LLMs (large language models), such as hallucinations, incomplete responses and inconsistent responses. For example, a request to summarise a case may return a complete, detailed response at one moment and then respond with an incomplete summary to an identical request,” he added.

“The cost of utilising AI systems also makes it unrealistic to use them across the business. The current processing power required to produce responses and the licensing models of vendors means that law firms will only really be able to afford one general AI tool (at most) and are more likely to deploy them to small sets of lawyers and staff. The firms that are currently using generic AI tools and those that are training their staff in prompt engineering are the ones that will be ready to utilise the next wave of legal AI tools most effectively.”

Automation and AI tools have been commonly utilised as efficiency tools – but now legal clients are also adopting them, something Chien described as a “more recent phenomenon”.

“From a client perspective, there is a palpable nervousness about confidential or proprietary information being breached or misused, especially in light of recent data breaches. Getting the balance right, as between achieving high-efficiency through use of tools and the risk of using these tools without any ‘lawyer’ input,” he said.

“It is incumbent on firms (and everyone in professional services) to demonstrate that the utilisation of these tools is not a means to replace or diminish the client service experience. It is also incumbent on them to understand, by talking to people like me, some of the legal issues with the use of AI.”

Similarly, there are a multitude of opportunities available for firms to both improve efficiency and their client service delivery, and these opportunities will only grow as new tech becomes available.

“On the legal side, there are significant opportunities for firms to use GenAI to help their lawyers collate large volumes of material, to identify key information or insights in one document or in a large suite of documents, and to increase the speed of legal research and drafting,” Sandler added.

“From a client service perspective, there are opportunities for firms to utilise GenAI tools and systems to share legal knowledge and resources with clients (and internally) in new and useful ways, to streamline communications with clients, to deliver legal services in ways that are more client-specific, and to equip legal staff with tools and technologies that allow them to spend less time on routine tasks and more time interacting with clients.”

As such, educating others as this tech evolves is key, as well as staying ahead of any extra developing risks, Howe emphasised.

“While we remain cautious as to its risks, AI has and will continue to streamline processes and enhance efficiencies. By leveraging AI-powered tools for tasks such as contract review, legal research, and due diligence, we’ve reduced the time and resources required for these tasks,” he said.

“It’s vital to always ensure the ethical and responsible use of AI, particularly in sensitive legal matters where accuracy and fairness are paramount. Firms need to navigate issues such as algorithmic bias and data privacy to ensure and maintain client trust and uphold professional standards. That’s why educating our people and clients on the legal risks associated with the technology is one of our priorities, so our clients can navigate the potential challenges of AI as it evolves.”

AI and a firm’s business strategy moving forward

The key focus for many BigLaw firms moving through the second half of 2024 will be to adopt and integrate GenAI into their existing processes, if they haven’t done so already. This will be a business imperative for firms looking to streamline legal processes, enhance efficiency, and deliver more value to clients.

“The successful firms will be the ones that do not just implement GenAI into their business, do not just use it to create new opportunities, greater efficiencies, and improved client services, but do all of this while still accounting for the risks and challenges posed by GenAI. Our advice for firms embarking on their GenAI journey is to make sure they have set an AI strategy that appropriately accounts for those opportunities as well as those risks,” Sandler advised.

“This includes consideration of the safe and responsible use of AI and an understanding of the costs required to introduce, test, implement and maintain AI technologies. In addition, it is imperative that firms understand the use cases for AI from the outset, and only then bring on the appropriate technology for those use cases. We have never found success in bringing in technology for the technology’s sake.”

Trialling a range of products is also something McLean recommended for firms that haven’t yet explored the realm of AI – and he compared GenAI to that of word processors and the internet in their early stages.

“As the AI legal market is in a very early stage of maturity, I would recommend that firms trial a broad range of products rather than jumping all in with a particular AI tool. The market is rapidly evolving, as seen by the announcement of a new legal AI product almost every week and the release of new LLM models such as Claude 3 last month and Llama 3 this week,” he added.

“The main piece of advice I would give to law firms is to expose as many staff as possible to generative AI tools and assist these staff to develop strong skills in prompt engineering and iterative content generation.”

But while tech skills are important to develop, the rise of AI tech will also mean that the “human skills” of lawyers will become more important, including designing solutions, abstract reasoning, critical review and analysis, strategic and commercial judgement, and understanding emotions, Howe explained.

“Lawyers bring important skills and experience [that] no AI can currently match, including an understanding of our clients and their people. We see AI as helping with time-consuming tasks and improving quality and efficiency if used and supervised correctly, but not as a replacement for the common sense, judgement, and wisdom that an individual can bring,” he said.

“Technology literacy has always been a key skill for legal professionals, and this now includes AI. Lawyers will also need to continue to adapt in how they apply the law to new situations where technology is involved, as it has an impact on many areas of law, such as governance, compliance, intellectual property, discrimination, privacy, workplace relations and contracts. Law reform will continue to occur to adapt to new issues, and lawyers will need to continue upskilling in relevant areas.”

Moreover, the role of lawyers will continue to evolve as a shift into more tech-led working practices allows law firms to thrive in a rapidly changing and digital landscape.

“This is an exciting time for the legal profession. The emergence of GenAI means that there are many new opportunities for legal technology to support lawyers in their service delivery to clients and for firms to gain strategic and competitive advantages. It also provides an opportunity to collaborate closely with our clients on our use of GenAI.

“Making the most of these opportunities requires engagement from everyone within the firm. Our firm has responded very positively to the potential of GenAI systems. Gilbert + Tobin has actively encouraged all our staff to learn about the potential of GenAI, to share their thoughts on how they see it impacting upon their work, and to test and provide feedback on GenAI systems. GenAI will impact our firm and everyone that is a part of it, so it is essential that everyone be part of our GenAI journey,” Sandler concluded.

“For the global legal profession, it will become increasingly important for the profession to engage and upskill with new technology as it becomes available. The lawyers that choose not to engage with these technologies will inevitably be the ones that will be left behind.”