AI's input revolution is fascinating: Here’s why lawyers should pay attention

Remember command lines? For those of us who navigated the early days of computing, the fear of a misplaced character was real. It was a digital language demanding precision and, frankly, a touch of masochism: one wrong command, and you could be staring at a blinking cursor wondering where all your data went. Things got better on the user interface front with graphical windows, mouse clicks, ribbon overlays, and swiping, but we have all become accustomed to interacting with computing devices in a certain way.

Today, the paradigm is shifting more rapidly than ever. Thanks to initial developments with voice dictation and assistants, and now advanced Artificial Intelligence (AI)-powered speech technologies, we are entering an era of natural language interfaces. As one group put it, “What rapidly transformed AI from a specialist technology into an everyday tool was to great extent a revolution in UI & UX.” The new generation of complex and revolutionary technology has become easy to use for all.

While many legal and scholarly discussions focus on perceived harms from AI-generated content (a controversy that will always capture headlines and spark debates), the real, immediate transformation for professionals lies in the power of natural language input. Professionals can now communicate with computers and achieve meaningful results for complex tasks, even when using somewhat imperfect, everyday speech with grammar or spelling mistakes, pauses, or filler words or sounds. Need to summarize documents, create tables, conduct research, or proofread correspondence? It’s becoming remarkably straightforward, like having a highly efficient (if slightly predictable and robotic) assistant readily available.

This is not just about the flashy output of AI image generators or GPT novelists. It’s about a fundamental, and often overlooked, shift in how we input information and instructions to our technology. For those who have sidelined AI as something they won’t (or can’t) bother with, or for those concerned about AI replacing your expertise, we can all relate to managing workload, meeting deadlines, and staying ahead in an increasingly demanding professional landscape. And it is there where the AI input revolution offers a practical, and frankly, essential advantage when carefully, and professionally, applied. 

This article is a follow up to our original article, Using AI responsibly as in-house counsel: Law Society of BC releases guidance on professional ‎responsibilities.  The Law Society of British Columbia, in its guidance on professional responsibility and AI, emphasizes the importance of lawyers understanding and responsibly using AI tools. As stated in their resource, "Lawyers have a professional responsibility to understand the benefits and risks associated with relevant technology”. This isn't merely about keeping up with trends; it's a core aspect of professional competence.

Natural language advantages

Legal technologies have been replete with arcane (but useful!) search syntax and user interfaces seemingly designed to frustrate the uninitiated. But we are rapidly approaching the reality that lawyers will be able to simply ask a computer to perform tasks in plain English. There are a lot of implications to the legal workforce behind the idea of simply asking a machine to "summarize this contract and flag clauses related to liability”, “find recent case law in British Columbia concerning privacy breaches in healthcare”, or “tabulate these financial records and highlight any unusual patterns”.  Set aside those longer-term impacts and think purely about the way it changes how a lawyer might interact with the machine that dominates their desk. This isn't science fiction, these are capabilities readily available today.

This shift to natural language input is significant because it offers:

• Enhanced accessibility: no specialized coding knowledge is required, lawyers can leverage sophisticated computing power using their natural language skills in a user-centric way;
• Improved efficiency: lawyers can translate thoughts into actions with unprecedented speed, shifting the focus to the substance of work, not the mechanics of interacting with technology; and
• Reduced cognitive burden: it offloads routine and repetitive tasks to AI assistants, preserving cognitive resources for higher-level analysis, strategic thinking, and client interaction.

E-discovery, for instance, is already widely using AI to help lawyers scan large swaths of electronic information for relevant documents. Current e-discovery software allows lawyers to scan documents using search terms or specific parameters, providing almost instant responses that are significantly faster than manual review. However, imagine the efficiency gains when lawyers can simply have a conversation with their e-discovery system, refining searches through natural dialogue rather than carefully constructed Boolean queries. "I need documents from 2023 where Johnson discussed the merger with Smith, but only internal communications" could replace complex search syntax, making sophisticated searches accessible to all practitioners regardless of technical expertise. The time saved and certainty available is only increasing.

Similarly, legal research — a cornerstone of legal practice — is being transformed by AI-powered tools that allow legal professionals to quickly scan and search large databases of regulations, statutes, practice areas, jurisdictions, and case law. The next evolution will be truly conversational research assistants that can understand nuanced legal questions, identify relevant precedents, and explain their reasoning in plain language. Rather than forcing lawyers to translate their legal questions into database queries, these systems will accept questions in the same form a lawyer might ask a colleague.

The implications extend far beyond research and discovery. Contract review, document drafting, case analysis, and client communication are all areas where natural language interfaces are beginning to demonstrate their value. By allowing lawyers to express their requirements in natural language, these interfaces reduce the technical barriers to leveraging computational power for legal tasks. This democratization of access is particularly important for small firms and solo practitioners who may lack dedicated technical support staff.

Developing a good prompt

Harnessing the power of AI input isn't about expecting magic; it's about mastering the art of effective instructions, known in AI as prompting. Indeed, prompting AI is remarkably similar to delegating a task to a bright but inexperienced junior lawyer or articled student. It would be unrealistic to simply instruct them, "go research contracts”, and expect a polished memo. Clear instructions, a defined scope, explicit expectations, and necessary context are all essential for effective delegation, whether to a human or an AI.

Developing the skill of AI prompting is the key competence for proper AI use. Just as vague instructions to a junior delegate inevitably lead to irrelevant or misleading outputs, poorly formulated AI prompts will yield similarly unsatisfactory results. Conversely, well-crafted prompts, rich in detail and context, can unlock remarkable efficiency and insightful results. Mastering this skill is not just about getting the most out of AI tools; it's about developing a more precise and effective approach to instruction and delegation in general.

Like that junior delegate, AI (despite its immense processing power and data access that vastly outstrips any human capability) requires clear and precise instructions to be truly effective. The same principles of clear instruction apply to both human and AI delegates, but understanding why context is so crucial for AI is key to effective prompting. Consider the inherent context a student lawyer brings to any task. They: are operating in a professional setting; possess some legal knowledge or context from law school; understand jurisdictional context in that the task is being given in or about a particular place; are aware of the idea that law firms are client-focused; have some idea of basic professional norms; and, in the very least, understand the physical and mental properties of being called into a senior lawyer’s office and asked to take on a task.

AI, in stark contrast, lacks this inherent understanding — or, really, any understanding of the world around it. It operates solely on the information explicitly provided in the prompt. AI has no pre-existing knowledge of legal principles, jurisdictional nuances, or professional purpose any more than it has pre-existing knowledge of the works of Shakespeare or vast reams of social media posts. Unless it is directly instructed, it is inherently drawing from a much broader context than any human would.

Therefore, when using AI tools, crafting effective prompts means consciously and explicitly providing the context that a human delegate would implicitly understand. A good prompt will feature at least a few key elements (and I’m sure you will find these would result in clearer instructions to that delegated lawyer, too):

• Clarity and specificity: clear and specific prompts help LLMs understand the task at hand, reducing ambiguity that can lead to irrelevant responses. For instance, instead of asking, "Tell me about renewable energy”, a more specific prompt would be, "Provide a 200-word summary on the benefits of solar and wind power".
• Context: providing relevant context enhances the AI's understanding of the scenario, helping it generate more tailored responses. Contextual details might include the user's background, the desired outcome, and the environment in which the task operates. For example, specifying that you are a lawyer in a particular practice area or jurisdiction adds valuable context that the AI will take into account in its output, as would providing the reason why the task is needed, and the factual background in which the task is given.
• Exemplars and structure: AI’s absolutely love pattern-matching, so including examples or templates in your prompts can significantly improve output quality. By showing what you expect, you guide the AI toward producing more relevant and sophisticated responses. Similarly, structuring prompts using bullet points or numbered lists can clarify instructions and improve focus in the model's output. This organization helps LLMs process requests more effectively, especially for complex tasks that require multiple steps or components.
• Open-ended vs closed questions: using open-ended questions encourages deeper exploration of topics and richer responses, but may not result in a clear answer. Similarly, asking for a yes/no response might yield an answer (which may or may not be correct!), but may limit the output. As such, when using AI it is typically better to frame inquiries in a way that invites detailed analysis and yield more comprehensive insights.
• Persona-based customization: while this would be implicit to a human delegate, telling the AI what its role is can be critical to consistent results and assigning roles or personas to the AI can lead to more tailored responses. For example, instructing the AI to respond as if it were a marketing expert when working on a business pitch can enhance relevance and presentation in its output.
• Iterative refinement: just like your junior delegate may not hit the result correctly on the first try, prompt engineering is often an iterative process. Testing different phrasing and formats can help identify what works best for specific tasks. Regularly refining prompts based on feedback allows users to optimize their interactions with AI.

Practical applications across professions: Moving beyond theory to workflow

Embracing AI input tools is not just about efficiency, it's about fulfilling our professional duty to provide competent and ethical service in a rapidly evolving technological landscape. Ignoring these tools is not only inefficient but could also be seen as failing to meet the evolving standards of professional practice.

Protecting your clients first and foremost

While the accessibility and power of cloud-based AI tools are undeniable, professionals, especially those in regulated fields like law, must be acutely aware of the data management and client confidentiality implications. Entrusting sensitive client information to third-party cloud services introduces inherent risks. Data may be stored on servers located in different jurisdictions, potentially subject to varying legal frameworks and access by external parties. Furthermore, the terms of service for many cloud AI platforms may grant the provider certain rights to use or aggregate data, even if anonymized, which could raise ethical concerns — or, even worse, leak inputted “secrets” to the public either directly or via letting the AI train on your data. The onus remains squarely on the lawyer (not any tool used by the lawyer) to ensure client confidentiality is rigorously protected. This requires conducting thorough due diligence on AI providers, carefully reviewing their data security policies and terms of service, and implementing robust internal protocols for data handling when using these tools. Strategies such as anonymizing data where possible, using end-to-end encryption, and establishing clear firm policies on AI usage are essential to mitigate risks and uphold the fundamental ethical obligation of client confidentiality in the age of cloud-based AI.

The BC Law Society's guidance on professional responsibility and AI explicitly underscores this point, reminding lawyers of their duty to maintain client confidentiality when using technology. This includes understanding where client data is stored, how it is processed, and who has access to it when using AI tools. Beyond simply relying on the promises of AI providers, proactive measures are necessary. This might involve seeking assurances regarding data residency, encryption practices, and compliance with relevant privacy legislation like BC’s Personal Information Privacy Act. In some cases, exploring on-premises AI solutions or carefully vetting cloud providers who offer enhanced data privacy features might be warranted, particularly when dealing with highly sensitive client information. This could mean supporting clients with new obligations under Ontario’s Bill 194. In addition, lawyers should consider, and support clients in consideration of the guidance from the Ontario Human Rights Commission. Lawyers should also be aware of any requirements on tool use or data storage from their clients over and above professional responsibility obligations. Ultimately, responsible adoption of AI requires a balanced approach, harnessing the benefits while diligently safeguarding client confidentiality and adhering to the highest ethical standards of the profession.

Enhanced service delivery and professional sustainability

Clients stand to benefit significantly from the efficiencies gained through AI input tools. Reduced turnaround times, more comprehensive analysis, and potentially lower costs are all direct benefits of leveraging these technologies. As the Law Society guidance implicitly suggests, utilizing tools that enhance efficiency can contribute to better client service and potentially improve access to justice by making legal services more affordable.

Furthermore, in an era of increasing workloads and professional burnout, AI input tools offer a pathway to greater professional sustainability. By automating routine tasks and streamlining workflows, these tools can alleviate pressure, allowing professionals to focus on higher-level, strategic work and improve work-life balance. As the Law Society guidance encourages responsible adoption of technology, this includes considering the well-being of practitioners and creating more sustainable practices.

The practical benefits can be seen throughout the workday. For instance, AI tools can help capture and organize ideas during commutes or between meetings, automatically transcribe and summarize meetings to eliminate manual note-taking, assist with document creation, summarize email chains, and streamline workflow through task automation. These capabilities allow lawyers to focus their time and mental energy on the aspects of practice that require human judgment, creativity, and interpersonal skills — generally the most rewarding aspects of legal work as well as the outputs often considered more valuable to clients.

Reframing the AI conversation: From replacement to enhanced capability

Focusing on AI as an interface, rather than solely as a content generator, shifts the narrative from replacement to augmentation. It's about collaboration between human expertise and machine capability. This perspective is particularly crucial for professions deeply rooted in specialized knowledge and judgment. Lawyers, accountants, and other professionals can embrace AI as a tool to enhance their abilities, not as a threat to their roles.

The BC Law Society guidance implicitly supports this perspective by emphasizing the lawyer's ongoing responsibility for oversight and judgment, even when using AI. AI tools are seen as aids, not replacements for professional expertise. The human professional remains the ultimate arbiter of quality, accuracy, and ethical considerations. And the human professional remains accountable to the client for the work delivered.

Practical adoption strategies: starting small and building confidence

Adopting AI input tools doesn't require a wholesale, immediate transformation. Professionals can begin with targeted applications addressing specific pain points. For instance, a lawyer might start by using AI-powered tools for meeting transcription and summarization, generating ideas for a first draft, or helping with business pitches or external presentations. Or they could experiment with conversational interfaces for discrete legal research tasks before integrating them into broader projects.

The learning curve for natural language interfaces is generally less steep than for traditional software, as the primary mode of interaction is already familiar. However, as the Law Society guidance implies, lawyers need to develop an understanding of how to effectively utilize these tools and critically evaluate their outputs. And, as we have explained above, simply because the interface is more intuitive, continuous learning and adaptation are key when using AI tools that are still changing rapidly.

Embracing the interface revolution responsibly

The true transformative potential of AI for professionals may lie not in its content generation capabilities, but in its power to revolutionize human-computer interaction. Natural language interfaces are democratizing access to computational power and reshaping professional workflows across industries.

For legal professionals in British Columbia and beyond, embracing this interface revolution is not just an option, but increasingly an imperative — but one that we might, actually, enjoy and benefit from. While there is still a lot of work to do as we grapple with the full implications of what AI can and cannot do, the immediate future isn't about humans being replaced by AI, it's about professionals empowered by AI-driven interfaces that make complex tasks as intuitive as a conversation with a colleague.