Lawyers must possess many skills – legal knowledge, keen time management skills, persuasion and a gift for eloquence. But their technology stack is usually only reflected in Microsoft Word.

Our software driven world is bringing revolutionary changes to the legal industry and producing the so-called “legal engineers”.

Andrew Burt, a graduate of Yale Law School, is the Chief Privacy Officer and legal engineer of immuta. He is at the forefront of this growing field, which is bringing artificial intelligence to law firms around the world. Immuta is one of several companies committed to automated data management. The so-called automated data management involves the rules of who can access data and how to access data, including the definition of allowing users with any technical level, and the automatic implementation of complex security and privacy controls on data.

Burt’s first attempt in this field was to make the legal Memorandum (which raises a specific legal issue, or determines which laws apply, and how these laws should be interpreted) enforceable by the machine. Legal memoranda are usually heavy work done by paralegals and junior lawyers, whose task is to find applicable case law and precedents when dealing with new issues.

“Traditionally, lawyers write memos and debate orally,” Burt said. “As legal engineers, our job is not to write memos. Our job is to transform the burden of such laws. For immuta, in particular, to transform the legal burden of data (legal provisions or internal policies related to the use of data) into technology. That is, to turn our legal expertise into software.”

Immuta’s team of legal engineers helps customers turn piles of legal memoranda into automated data strategies and develop new product features. For example, Burt is considering designing a customer questionnaire for the health insurance portability and Accountability Act (HIPAA) to automatically generate data policy rules in immuta.

More generally, the role of a legal engineer is to establish an interface between legal and technical professionals in order to build software to interpret, enforce, or prove compliance with the law.

Some legal engineers are developers or data scientists interested in legal procedures. More commonly, they are lawyers with technical skills, which is still a very unusual combination in the legal industry. They are willing to automate some of the work they did when they first received training.

“As people with professional knowledge, our goal should be to ask ‘what is basically replicable?'” Burt said. “Let’s turn it into technology in an extensible way, and then let’s focus on what’s really interesting and why we become lawyers, which are often marginal cases.”

Burt began his legal career in the government. He is a special policy adviser to the assistant director of the network Department of the Federal Bureau of investigation, and is also the main author of the report after the 2014 FBI’s destructive cyber attack against Sony. Hackers stole a large amount of data and released employee data, internal emails, and even unpublished online movies. Burt handled the highly sensitive data in this serious network intrusion and the legal burden brought by these data. That means drowning in legal memoranda.

“We were just overwhelmed,” Burt said. “If the scale of the memo is not large, lawyers will want to write a memo. So, I basically fell in love with the idea that if the lawyer sends something that can be executed by the machine instead of writing a memo? What if we can automatically ensure that everything we do complies with the law?”

Burt taught himself the prolog programming language because it can be used to represent a logic based rule that defines the data charter. He found that an active research group had begun to solve the problem of how to express contracts programmatically. Eventually, he made contact with the founder of immuta, who wanted to eliminate data islands in large organizations.

“Once you solve the problem of data islands, you will find that you have encountered a legal problem, that is, there are rules related to data.” “So I told people that I was going to set up a legal engineering team at this startup, and they would say, ‘why? What is this?’ I clearly remember a colleague rolling his eyes at me,” Burt said

Roland Vogl is the executive director of Codex, also known as the Stanford legal information center. He was first interested in establishing a system to solve legal problems when he was a teacher at Stanford University, and later joined Codex in 2008. The center conducts research in the fields of computational methods, automation and mechanization of legal analysis. One of its core projects is related to computable contracts.

“From customization to systematization and standardization, there is always such a scope,” Vogl also said. “This is the entry point for legal engineers to help figure out what can be automated. In the automated workflow, how far can we upgrade our customers? And when do we need to upgrade them to manual decision makers?”

Codex has experienced an explosive growth in the activities of start-ups in the field of legal technology. The center’s legaltech index includes nearly 1200 early-stage companies, many of which decompose and automate specific legal tasks or processes. Sometimes lawyers and bar associations are annoyed by this. They have sued some technology providers for unauthorized legal practices.

For example, lawgeex uses artificial intelligence to automatically review contracts and mark non-conforming and missing terms. Traditionally, companies take a long time to review a new contract and decide whether it meets the standards in all respects. Vogl said: “most of this knowledge is hidden in the drafts of other previous negotiations, or in the minds of some senior in-house lawyers.” In a recent study, lawgeex achieved 94% accuracy in NDA, while the average accuracy of experienced lawyers was 85%.

Flightright in Germany has automated legal procedures to obtain compensation from airlines for delay or cancellation of flights. This service is free and has been used by 5.2 million people, with a 99% success rate in court cases.

“This has opened up a market that lawyers traditionally do not pursue, because it is a small number and is consumer oriented.” Vogl also said, “every case they lose will cost them, so they use advanced machine learning technology and data science technology to evaluate whether a particular case is a good case.”

Startups like this, as well as law firms that want to turn some of their services into products, will be among the first companies to need legal engineers. However, the legal industry as a whole lags behind other industries in the adoption of automation technology.

Vogl said: “law firms have seen this situation. This technology will really change the way legal services are provided.”

Codex also helps train lawyers in their future legal roles. “About 80% of people need to use technology more maturely, and we need to help them early in their education, perhaps in the first year of the course,” Vogl said. “Another 20% are innovators themselves. They can build new systems, and we need to provide them with courses to help them develop their ideas.”

Other law programs around the world are also preparing students for their future legal roles. The center for legal technology and innovation at Michigan State University offers courses in artificial intelligence and law, automated vehicles and law. Ie law school in Madrid offers a master’s degree in legal technology to lawyers who want to digitize the legal department of a company, or who want to create a technical department in a law firm.

Despite some exaggerations about robot lawyers, Vogl does not think that human lawyers will disappear soon. “As long as there are humans, there will be transactions and disputes. We need humans to solve these problems,” he said. “But these machines can help us get rid of monotony and deal with low-level things, so we can focus on legal judgment, which is what lawyers and legal professionals are good at.”

Responsible editor: CT


Leave a Reply

Your email address will not be published.