top of page

No, Robots Won’t Replace Lawyers

Evan’s Data Corp recently surveyed 550 software developers, asking them what was the most worrisome aspect within their careers. A resounding 29% of those surveyed responded with the answer that they believe their jobs will be replaced by artificial intelligence. It a common fear in the job market, but even attorneys are starting to feel the pressure. The 2015 Altman Weil, Law Firms in Transition Survey showed that only 20% of managing partners believed that computers would never replace human practitioners – that leaves 80% that are, at the very least, not sure if human lawyers are irreplaceable.

The growing popularity of legal technologies, assisted review, and automated practices has given the legal industry what we’re calling “automation anxiety,” a nervousness about technology automating the work lawyers do – and with good reason. There’s an estimated 13% of routine review task work that could be replaced by computers. If that work was replaced by computers within the next year, there would be a massive job crisis. But what the more optimistic futurists argue is that there’s no way computers will be ready to replace that work overnight. It will take several years before the technology is advanced enough, and siphoning out 13% of work over several years will hardly be missed. What’s more, computers and robots will arrive to enhance, not replace our jobs. We’re looking towards a future where the human workplace is transformed, not eliminated. In fact, in 2015, 149 million people were employed. That’s the highest number in history.

David Autor is an Economics professor at MIT specializing in the effects of automation on employment, and he assures that while, yes, technologies have cut into the market, it’s been in a much more targeted way than we typically discuss. He argues that machines are here to help make us more productive, not to phase us out. What he says is being eliminated is the “task work,” not the job. When menial, mid-level task work – like clerical duties – can be eliminated, we see more time for employees to spend on the work that counts, and it increases the demand for jobs that require critical thinking and abstract reasoning to control those machines. What we’re seeing are mid-level jobs being cut, like secretaries and administrative positions, but a rise in both low-level and high-level jobs. Until computers have the ability to intuit nuance the same way the brain can, there’s no way our jobs can be replaced. Many people understand parts of their job without being able to explicitly explain what or why they’re doing it. If a person can’t describe what they do, a program cannot be written to automate it.

In the legal industry, what will protect jobs is the highly nuanced, highly variable nature of the profession. The work of attorneys involves a wide range of tasks, and the part that is most easily automated – reading and analyzing documents – only amounts to a fraction of the work. Counseling, court hearings, depositions, litigation, and other more complex work performed by lawyers is too variable to automate. A lot of what is required is based on the possibility of infinite outcomes. It would be impossible to write a program that could account for all the contingencies in play.

One prime example of how attorneys are already benefiting from artificial intelligence and technology is in research. Technologies like Westlaw, LexisNexis, and even Google have given attorneys the ability to instantaneously search millions of legal resources, trial records, etc. for the information they need. These technologies are so ingrained in attorneys’ lives and workflows that, at this point, past research methods seem like poring through dusty law books by lamplight. Then there are the tools that automatically organize, refine, and relate data during discovery. In the past, there’d be mountains of file boxes to sort through, but now there are e-discovery technologies that can cull data, remove duplicates, and cluster documents into concepts and families so that attorneys don’t have to worry about the noise; they can focus their expensive review time on the documents that are likely to be relevant to their case. Even further into the e-discovery workflow are tools that use automation and predictive coding to relieve lawyers from the mundane and time-consuming efforts of more pinpointed review practices – for example, our tool Blackout automates the redaction process to save attorneys from the tedious manual redaction process.

At first glance, these legal technologies might sound like they’re automating the work of lawyers in order to replace them, but really they’re freeing lawyers from tedious review tasks and freeing them up for the more thoughtful, nuanced work that makes them irreplaceable. In an article for the Wall Street Journal, Thomas Davenport (Research Fellow at MIT) succinctly describes why the growing automation anxiety in the legal profession is counterproductive; rather than worrying about how technology might replace them, “Smart lawyers should learn what these technologies can do and use them to augment their own work.”



bottom of page