New Zealand Law Society

Navigation menu

Developments

29 March 2018

Lawyers vs AI

LawGeex is a software company based in Tel Aviv, Israel. Early in 2017 it released software that uses artificial intelligence to study contracts, pinpointing any language or requirements that seem unusual.

The company’s Contract Review Automation software is marketed as the quickest and easiest way for businesses to review and approve incoming contracts. LawGeex claims to save 80% of the time required to review and approve, and 90% of the cost when compared to manual review. It has focused its marketing on in-house teams – not as a substitute for lawyers, but as a time-saving tool for the “time intensive and mundane task” of reviewing contracts.

An illustration of light bulbs and machines

In February the company released the results of a recent study it had carried out with independent auditing by legal academics from Stanford University, Duke University and the University of Southern California. The study involved 20 lawyers with experience of reviewing contracts at large companies and global law firms. They were matched against a LawGeex AI platform which had been trained to evaluate legal contracts. The training had involved tens of thousands of contracts using machine learning and deep learning technologies.

The study involved five non-disclosure agreements, with participants given four hours to review them and to identify 30 legal issues, including arbitration, confidentiality of relationship, and indemnification. The participants were given a score for accuracy of issue identification.

The result showed that the human lawyers averaged 85% accuracy – against 95% accuracy attained by the AI. And, more tellingly, the AI finished its task in 26 seconds. The humans averaged 92 minutes.

LawGeex understandably ensured the results were well publicised, including the following comment from University of Southern California Professor of Law and Economics Gillian K Hadfield:

“This experiment may actually understate the gain from AI in the legal profession. The lawyers who reviewed these documents were fully focused on the task: it didn’t sink to the bottom of a to-do list, it didn’t get rushed through while waiting for a plane or with one eye on the clock to get out the door to pick up the kids. The margin of efficiency is likely to be even greater than the results shown here.

“This research shows technology can help solve two problems – both making contract management faster and more reliable, and freeing up resources so legal departments can focus on building the quality of their human legal teams.”

“We’re not claiming to be more accurate than lawyers all the time and in any type of work,” LawGeex CEO Noory Bechor says. “What we are showing is that on the mundane, repetitive, simple stuff, technology can actually do a better job that humans.”

More details: The full study can be downloaded from AI vs Lawyers

Techlaw intern programme

Simmonds Stewart and the University of Auckland Law School have launched a techlaw intern programme. This offers law students paid intern positions at companies in the high-growth tech centre.

The aim is to provide law students interested in technology and entrepreneurship real-world experience in the tech sector. Each student will be paid up to 100 hours of basic legal work for a tech company. Initially the programme offers internships to between five and eight University of Auckland law students.

Participating companies include Jude, Endace, 9 Spokes, Pacific Channel and Auckland UniServices. Simmonds Stewart will provide training to participating students, as well as pairing the student with a Simmonds Stewart buddy who will support the student and supervise their work where required. Penultimate and final-year students may apply. Applications for the first intake closed on 13 March 2018.

More details: Simmonds Stewart

New NZ legal technology organisation brings vendors and users together

Newly-formed organisation LegalTechNZ aims to bring together people with a shared interest in improving the use of technology in the legal sector. Its stated vision is “a thriving New Zealand legal sector which understands and applies technology to improve the efficiency of and access to legal services in New Zealand”.

Among its membership, LegalTechNZ includes several prominent law firms, InternetNZ, Spark, Kōwhiri, TrustUs and Dacreed.

The organisation’s chair, Simon Stockdale, is CEO of New Zealand contract review software company Pagemap.

“As vendors we often see substantial changes in how law is delivered in other markets through the technological advances, sometimes using software developed in New Zealand,” he says.

“Vendors want to find a better way to engage with the New Zealand legal profession and help it transition to a digital future, for the benefit of the legal sector, its clients and the New Zealand legal tech providers.”

Hudson Gavin Martin partner Tim Mahood says his firm became involved in establishing LegalTechNZ because it wanted to help ensure that the industry “is prepared for the change rather than having it forced on them”.

“As a Technology Media and IP law firm we see every day how technology is impacting the way that business is carried out. Law is not immune to this,” says Mr Mahood.

“We see that technology, when used right, can improve the way lawyers provide services to their clients. We recognise that there is a risk lawyers will want to carry on doing what they currently do – but as in other industries the customer will drive the change whether or not lawyers want it.”

More details: Legal Tech NZ

“Legal AI for the #MeToo era”

Legal AI startups are now almost a daily event with many targeting the larger, more lucrative end of the market. One of the latest, Attorney IO was launched in February and says it was founded on the premise that legal AI can finally fulfil the guarantee that everybody deserves access to true justice: “The goal is to empower David to beat Goliath”.

The United States-based startup describes itself as “legal AI for the #MeToo era”. “The first generation of legal AI focused primarily on helping the biggest firms represent huge corporations and the 1%. People like Harvey Weinstein hire some of the richest lawyers … to obscure abuses. Many of these lawyers then boast about using an army of AI and human lawyers to stomp on victims and ensure they don’t get a fair day in court. We want to finally empower the people most in need of justice.”

Attorney IO specialises in finding relevant case connections from “millions of legal cases in our collection” which it says are a low-cost opportunity for attorneys to get a second pair of eyes on their work product.

More details: Attorney IO

Some recent legal AI research

AI, legal ethics, smartphones and lawyers

Legal ethics and how they fit with new developing AI and other technological processes are starting to exercise the minds of the academic community. “The first part of this article defines the brave new world of AI and how it both directly and indirectly impacts the practice of law,” the authors state. “Part two explores legal ethics considerations when selecting and using AI vendors and virtual assistants. Part three outlines technology risks and potential solutions for lawyers who seek to embrace smart phone technology while complying with legal ethics obligations.” The authors also urge the legal profession – “known for coming late to the technology dance” to step in now to take control of AI’s impact on the profession.

More details: Jacobowitz, Jan L. and Ortiz, Justin, “Happy Birthday Siri! Dialing in Legal Ethics for Artificial Intelligence, Smart Phones, and Real Time Lawyers” (8 January 2018). Texas A&M University Journal of Property Law, 2018; University of Miami Legal Studies Research Paper No. 18-2.

Robots and their criminal liability

“If we are attacked by an intelligent robot, can we impose criminal liability upon robots? How can we defend ourselves legally?” asks Professor Gabriel Hallevy of the Faculty of Law, Ono Academic College. Posing the question: “Does the growing intelligence of AI robots subject them to legal social control, as any other legal entity?” Professor Hallevy attempts to work out a legal solution to the problem of the criminal liability of AI robots. He concludes that if all of its specific requirements are met, criminal liability may be imposed upon any entity – human, corporate or AI robot. “AI robots have no soul. Thus, there is no substantive legal difference between the idea of criminal liability imposed on corporations and on AI robots. It would be outrageous not to subordinate them to human laws, as corporations have been.”

More details: Hallevy, Prof. Gabriel, Dangerous Robots – Artificial Intelligence vs Human Intelligence (11 February 2018).

Using AI to hire people

What happens if a company decides to delegate its hiring decisions to a computer, instructing it to “Pick good employees”? Even if the computer could use all the employer’s available data and find anything else it needs on the internet, how would it deal with diversity?

More details: Sullivan, Charles A, Employing AI (18 February 2018). Seton Hall Public Law Research Paper.

Artificial Agents and General Principles of Law

Artificial agents challenge general principles of national and international law, says Professor Antje von Ungern-Sternberg of Germany’s Trier University. She proceeds to look at the principles of responsibility, explainability and autonomy and concludes that lawyers will have to define the areas of law that require an explanation for artificial agents’ activities. Artificial agents do not only challenge existing principles of law, they can also strengthen them as well, she concludes.

More details: von Ungern-Sternberg, Antje, Artificial Agents and General Principles of Law (28 January 2018). German Yearbook of International Law (Forthcoming).

Developments aims to provide information on new products, services, research and other information on the Future of Law which is likely to be of interest to New Zealand lawyers. The Law Society does not warrant or endorse any of the items included.

Last updated on the 29th March 2018