The future of law? How generative AI is set to change the legal landscape
Since February, the news has been dominated by the excitement and advantages of AI technology. With new advances in ChatGPT and the efficiencies it’s bringing to the workplace, it’s no surprise that the legal industry is looking at ways it can embrace the benefits.
The legal industry is slowly dipping its toe into generative AI solutions, with early adopters, such as Allen & Overy, making full use of the technology since February. A recent survey conducted by Thomson Reuters Institute examining generative AI in the legal industry has found that more than a third of respondents were deliberating whether to adopt generative AI at their firm. However, despite this intrigue, it’s not yet common across the sector, with only 3 per cent of respondents currently implementing solutions.
In this blog, we dig deeper into the possibilities of generative AI with our Operations Director, Owen Morris. Owen shares his insights on how generative AI operates, its utility, the challenges and opportunities it presents, and, most importantly, how it is poised to optimise the running of the legal industry for those who adopt it.
What is generative AI, and what makes it useful?
Generative AI is a powerful tool that uses machine learning to generate new content. There are several different types of Generative AI that can create content through image or text generation.
Much of the recent discussion has been about ChatGPT, its integration with Microsoft Bing and other text generation systems, such as Google Bard and the underlying Palm2 models announced at Google I/O. These systems run on Large Language Models (LLMs) and respond to prompts based on their training. Unfortunately, this makes them uncannily good at things that they know a lot about but less reliable when it comes to unique insights and creative responses.
In essence, using generative AI is like having a chainsaw instead of a handsaw. It can make tasks quicker and easier to complete. However, you’ve got to make sure you know what you’re doing with it. The responsibility for the content lies with the user, who must interpret the results and determine their usefulness and reliability. Otherwise, the content could be ineffective at best and prejudiced at worst.
Generative AI is a rapidly advancing field. ChatGPT from OpenAI currently leads the pack, yet other companies are striving to exceed and surpass its capabilities. This is an exciting space that is opening up a multitude of possibilities for the global workforce to take advantage of. Microsoft’s partnership with OpenAI is especially interesting, combining Microsoft’s data with OpenAI’s smarts in the Copilot product suite.
How can legal workers optimise their workflows with generative AI?
Where I see it making the most impact is the head start it’ll give firms by dramatically accelerating the speed of drafting documents – client bids, interview summaries etc. Most firms will already have templates for common tasks, but the true value of generative AI lies in swiftly crafting written responses for less ordinary tasks.
This advantage empowers firms to quickly produce responses that their team can analyse and enhance. It accelerates productivity, allowing teams to allocate their time wisely to tasks that truly contribute value rather than mundane everyday activities.
How could the use of generative AI change the way legal services are provided?
Right now, as legal firms start to implement solutions incorporating AI technology , the possibilities are endless. For example, firms can now quickly generate the foundations of deals and proposals – giving them a major advantage over competitors. Other good use cases are research, digital dictation and document assembly.
As the technology advances, these tools will become increasingly personalised, catering to the individual needs of each user. Microsoft’s latest release, Microsoft Copilot, is already allowing users to train models with their firm’s own data, providing a notable advantage for firms and enabling tailored responses specific to their use case.
In the long term, as more firms adopt these solutions, I expect these tools will become standard practice. This shift will require firms to work harder to differentiate themselves, likely prompting them to encourage their teams to embrace these tools as quickly as possible.
What tips and best practices should legal firms follow to get the most out of their generative AI implementations?
Starting out, the best advice is to try the tools out. Figure out how they work, and specifically what works and doesn’t work for you and your team. Piloting these solutions with specific practice areas is a good way to gain feedback and demystify the technology.
From there, it’s a case of tracking your day-to-day work. Are there some common tasks that AI can assist with or make more time-efficient? Outside of the usual daily tasks, AI can also serve as a valuable aid in areas where legal professionals may lack expertise.
For instance, ChatGPT can provide insights into complex matters through simple prompts – helping professionals get ahead. What Generative AI does is eliminate the relentless scrolling of Google searches and gives you an instant answer on a matter that you may be unfamiliar with.
What challenges and opportunities exist for legal professionals when adopting generative AI solutions?
With the good comes the bad. AI opens a lot of doors for legal professionals, such as generating human-like responses to complex questions. However, there’s no guarantee the model is producing an accurate answer. So, if you’re not already an expert in the field, it can be challenging to interpret whether the answers you’re getting are reliable or not.
With that in mind, although machine learning has been an integral part of many areas of the legal profession for specific applications, the integration of newer AI tools into the wider legal field should be approached cautiously. It should be viewed as a tool that complements legal professionals rather than replace them, emphasising human-machine collaboration. Human judgment remains essential in the partnership to assess the generated content’s suitability and accuracy.
While, in its current state, AI will not replace human teams, certain tasks can be better initiated by machines. This is not unique to generative AI and can be observed widely whenever new technology is introduced. Building AI into everyday processes is, therefore, likely to become a norm across industries.
A particular challenge will be ensuring the security of data entered into the models. Legal professionals must ensure that confidential data remains protected, avoiding prejudiced responses or compromises to their firm’s security and reputation. Implementing appropriate policies and security controls is key. There is also an opportunity to integrate AI enabled protection into cyber defences for safeguarding operations. Looking ahead, I expect we’ll see the legal world prioritise building secure foundations for generative AI, enabling leaders to confidently leverage its efficiency while mitigating current risks associated with the technology.