Our purpose

How AI and machine learning is shaping legal strategy

Isabelle Moulinier, VP, Applied AI Research

Five years ago, many experts predicted that we would routinely see self-driving cars on the road in 2021. That has not come to pass. What we do have are cars where Artificial Intelligence (AI) can assist drivers. Forward collision warning, lane departure warning, or rear drive assistance are all AI-enable features that make my life safer.

Why talk about cars in the context of AI and the law? Because it illustrates a shift from full automation to assistance, or augmentation – helping individuals perform some tasks better, faster, smarter. And augmentation, rather than automation, is key to the role AI and machine learning can play in shaping legal strategy.

When we think about AI today in legal tech, it is important to remember that AI is not merely one thing. Instead, it is a variety of technologies and task-specific applications that can assist legal professionals in the exercise of their function. (The future of law firms and lawyers in the age of artificial intelligence)

Today, the legal tech landscape is rich with AI solutions. There are AI tools designed to automate simple tasks like time keeping or validating billing entries, and there are AI tools designed to help legal professionals accomplish some tasks faster and more accurately.

Legal databases have put the law at everyone’s fingertips. At the same time, though it remains largely untapped, the amount of data generated within law firms keeps growing. Solutions leveraging AI and machine learning – a type of AI that finds patterns in a lot of data – can help legal practitioners gain the information and insights needed to prepare for litigation, draft documents and verify their work products. Technology is enabling fast, accurate research, and cutting down on the time and cost of legal work.

Increasingly, search providers offer tools such as semantic search (using new machine learning representations to find relevant documents) and passage level retrieval (pinpointing relevant portions in documents. For some time (two decades), e-discovery has used classification, a kind of machine learning, to automate the lengthy process of document review. Armed with a database of past negotiations or verdicts, public or firm-specific, lawyers can analyze past outcomes and use newly gained insights to assess the likely outcome of a specific scenario.  

Some AI solutions offer the ability to automate the review of legal research, allowing judges to check submissions and identify missing precedents and authorities. Similar tools are available to attorneys as well. Similarly, in the transaction space, AI can be used in document automation and leverages large databases of precedents to assist the drafting and analysis of documents.  All these innovations demonstrate the growing potential of AI within the legal space.  

One additional area where AI has potential to improve access to justice is in expertise automation. Some AI tools automate simple workflows in civil matters, providing wider access to the law to people who cannot afford legal counsel. Legal professionals may also benefit from expertise automation as a means to streamline common tasks and free up more time to think deeply about their clients’ cases.

Some of these tools include drafting a will. An offering like myopencourt.org “enables everyone to get the basic help they need to determine the strength of their legal disputes.” Similarly, there are apps available in the United Kingdom and the United States to help people dispute parking tickets. Such services often rely on “expert systems,” which embed specific legal knowledge in rules and decision trees that can include calculations or factor weighting, among other techniques.

It’s easy to think that technology, including AI, is fairer than humans. But humans create technology and AI, and human biases can seep into the design, development and use of AI systems. This is particularly challenging with machine learning, which relies on data that may capture societal inequalities. Regulation and standards for AI use are nascent in most industries, and it’s important for people who design, develop, and use AI tools to work to develop processes to identify and manage associated risks. And while enhanced scrutiny around data has the potential to make inequalities in historical data more visible, it also enables the development of mitigation strategies.

The use of AI in the legal profession has only just begun. While we have already seen its impact, its continued use and growth will shape legal strategy in the years to come. Just like the self-driving car, just because lawyers aren’t being replaced by machines, doesn’t mean that technology isn’t already making its mark on the legal profession. AI helps augment human expertise, it does not replace human expertise, and incremental innovation through AI will continue to combine with human intelligence to enhance the pursuit of justice.  

Learn more about 

Inside Thomson Reuters blog

Sign up to receive new posts by email