Although AI offers a lot of promise for the legal field, it’s not a golden ticket to bigger profits and improved efficiency. Before you take the leap and adopt AI for your organization, you should understand how it works: its processes, its capabilities, and its limitations. In this article, we’ll discuss five of the most important things that every general counsel should know about artificial intelligence.
1. AI Will Not Replace Humans
Although the jury’s still out on how exactly AI will impact the workforce, it’s clear that we won’t be living in a jobless dystopia where robots fill everyone’s position.
Some duties and tasks, many of them menial and repetitive, may be eliminated or reassigned. However, it’s possible that many existing roles will become more robust, and new functions and responsibilities (such as legal engineers and IT/technology lawyers) will appear. In fact, it seems likely that as AI and other technologies begin to permeate the legal industry, a new set of needs will also arise.
Law is inherently a human practice: it was created by human beings and intended to apply to human actions. As such, some aspects of legal practice will always need a certain human touch.
For example, humans still outperform even the smartest AIs when it comes to more subjective, higher-level activities like analytical thinking and responding to unpredictable events. In addition, as priorities shift, humans are better able to an evolving situation because they can see the larger picture and understand the final desired outcome. Humans are capable of looking at intangible qualities and teasing out the relevant details, in ways that computers cannot, to find better solutions to their problems. Finally, even the best the AI will struggle with ethics and judgment. The practice of law requires the ability to read between the lines, to interpret language, and to understand when to apply the rules and when the rules might change depending on a specific set of circumstances. Shades of gray are complicated, and often fail to permeate the cold, rational, programmable robot brain.
2. AI Makes You More Efficient
On the other hand, when it comes to AI and business productivity, the rumors are all true. Consulting firm McKinsey & Company estimates that 23 percent of a lawyer’s job can be automated, including document review and search, calendar alerts and deadline reminder management, and analytical reporting that can be customized to reveal specifics about an organization’s critical documents and contracts. AI can save you an astronomical number of man-hours by automating your repetitive tasks, such as searching for and tagging documents.
One of the most significant costs experienced by most legal departments involve humans - all those hours devoted to looking through documents, tracking down emails, making changes, even dotting those “i’s” and crossing those “t’s.” Saving all that time spent creating, researching, and managing contracts translate into dramatic cost savings. Automation and AI can cut through those tasks like a knife, reducing the time spent on traditional tasks, like document creation, review, and management, from days to minutes.
Every general counsel would love to have one-quarter of their lawyers’ time suddenly available. The time and efficiency savings from artificial intelligence permit you and your staff to concentrate on more meaningful, strategic activities for business growth.
3. The ROI of AI is Need-Dependent
When it comes to ROI, not all artificial intelligence is created equal.
It’s true that some AI solutions will be complex and expensive, but the good news for small teams, or legal departments with a limited budget, is that many AI options will be affordable and straightforward. For example, much of the AI currently being developed for contract management exists as an additional feature, rather than an entirely new software platform.
Determining the ROI of your AI solution can be tricky, and it depends on what your business is looking for, so it’s helpful to ask the following questions:
What is the level of customization required?
How many users do you have in your system?
What is the level of functionality and complexity of your processes?
What resources will you need in terms of staffing and time for implementation?
Fortunately, AI doesn’t have to be expensive, and it doesn’t have to be difficult to implement. Many AI tools are designed to work in concert with existing systems, amplifying and enhancing your current workflows rather than introducing new processes. Ultimately, the costs and benefits of AI are a function of your business goals and requirements.
4. AI Has Risks and Biased Outcomes
Even though AI is designed to be “objective,” seeing only the raw bits and bytes, it’s not immune to bias and unintentional discrimination.
For example, in 2014 Amazon built an artificial intelligence tool to automate part of the hiring process for one of its engineering teams. The team shut down the tool after discovering that it rejected women’s resumes at a higher rate. Because the resumes use to train the AI included a larger proportion of male candidates, the AI rated men’s resumes more highly when analyzing real-world examples.
In general, problems with incomplete and inaccurate datasets can plague AI solutions, delivering strange and unexpected results. AI also needs to be robust to a variety of situations that software developers may not be able to anticipate or program for in advance.
What’s more, because AI models are so complex, it can be difficult to determine precisely how the software reached a particular conclusion or took a specific action. Neural networks are one example of an AI model with millions of different parameters, making them impossible to analyze by hand.
5. AI is Evolving
One thing is clear: AI will significantly disrupt the modern legal department. What’s not yet clear, however, is how it will do so.
The precise effects of techniques such as advanced automation and predictive analytics are still being evaluated, and researchers are still discovering new capabilities and limitations. Because AI is such a fast-moving field, the landscape for legal AI may look very different even 5 or 10 years from now, in ways that are impossible to guess.
If you’re wondering whether you can use AI in your legal department right now, the answer is yes.
Straightforward, inexpensive AI software is already available for simple tasks such as electronic discovery and keyword search. In addition, more sophisticated solutions for responsibilities such as contract analysis, document review, and legal research have begun to roll out across forward-thinking legal departments.
However, you need to perform your due diligence before adopting any AI tool for your enterprise:
Stay up-to-date with the latest AI news and developments.
Thoroughly vet third-party vendors before choosing a technology partner.
Designate an AI “champion” or team responsible for discovering and assessing your different options.
Keep an eye on your company’s AI risk, compliance, and ethical obligations.
As the future of legal departments adapts to include new tools like AI, knowing what to look for and how to choose the right technology to fit your needs will reap tremendous benefits in terms of simplified workflows, improved organization, and increased productivity.