Not if but when: The rise and rise of AI in legal practice

Lawyers being replaced by technology, should we be scared?

Both legal professionals and aspiring ones have read about the recent developments of artificial intelligence (AI) in the tech industry, and have been amazed by fantasies of self-driving cars, patients undergoing interventions by surgeon-machines and programmes composing new musical hits all on their own.

For the less tech-savvy, AI is the process of software reading and analysing information, following up with answers, extracting information and refining their know-how, workflow and output quality.

Yet, what we legals have often overlooked is how such developments will affect our industry and our future jobs. What can programmes already do? Which tasks are already redundant? Here is a partial list of the names that should concern every legal professional, no matter their professional seniority:

• CARA (where the name stands for Case Analysis Research Assistant) will allow you to upload a document and have all the research on the legal cases mentioned therein in a blink of an eye. You can securely upload contracts, briefs, research papers and obtain all the results you need to fully grasp the legal panorama of your document.
• Kira Systems is a software already adopted by Clifford Chance that can search and analyse agreements and can even be trained. The law firm declared in this article that one of its priorities is to reduce its billing by having software undertaking actions otherwise assigned to costly humans. DLA Piper is another big client.
• Deftr can help you get the full picture of any corporate structure (‘who owns who’, ‘what’s its stake’ and more).
• DoNotPay: you may have already read about this software, which helps you appeal parking tickets. Naming itself the first “UK robot lawyer”, among its new purpose is to fight homelessness by informing people of their rights.
• Ross (used by law firms such as Dentons and BakerHostetler) is perhaps the scariest one: you can ask questions to this software in plain English and receive answers based on statutes. Perhaps raw answers, but still able to be redefined by its core code. ROSS is based on IBM’s Watson Technology, known as “cognitive computing”. Who’s next to follow?
• Beagle does something very similar: revising contracts. Do you need to know who are the parties, which liabilities they carry or any other specific information? Its sniffing method will help you out and save you time, for sure.
• Similar to this are LawGeex and Legal Robot: just upload your contract and have it compared and reviewed online.
• RAVN, instead, helps search for client names across regulatory bodies’ registers in Europe.
• RivierView Law’s Kim is the first virtual assistant that can support lawyers in their petty tasks and works on managing instructions and providing advice, among other tasks.
• PerfectIt and WordRake are online software you can use to proof-read your documents (be them contracts or something else).
• DocNav — although its website may be down — is an online document-viewer software that simplifies visualisation and editing of legal documents.

Beyond the names of these programmes, it looks like further boundaries are about to be broken. Here are some examples of developments in the pipeline:

• Algorithms that can help identify cyber-threats and could replace lots of security and compliance officers.
• Companies are already testing AI in their Board of Directors. Deep Knowledge (a Hong Kong-based company) is the first company to have appointed a software named Vital as a director; its vote will count as much as the votes of the rest of the firm.
• Virtual reality alone is being studied to help jurors reach a closer knowledge of crime scenes before issuing their verdicts.
• Premonition uses data-mining to check lawyers’ history and assess who has the better chance of winning a case.

Is there any safe harbour to this tech revolution? When could it possibly strike? What could be the possible consequences? How will governments react (and have they yet)? How does hourly-billed assistance cope with an instantaneous answers framework? What are the chances these programmes get hacked? Will a Cost Benefit Analysis still lean towards humans? Will this AI be regulated by or under any authority?

No one has a crystal ball to look into the future, but reputable voices have spent some words on these questions.

According to McKinsey, the most secure jobs are those that involve supervision on humans.

Senior consultants have warned it could take roughly five years before we see a massive deployment of these technologies, whose reach will stretch to senior lawyers too. Renowned futurologist Susskind uses the same range of time, citing the 2020s as a pivotal point.

Although many questions are left wide open, we can already trace a trend in the legal industry of the replacement of humans with machinery: ediscovery on the localisation of documents, due diligence on contractual terms and contract management/automation for the life-cycle of contracts.

On the other side, some are more cautious when it comes to the future of legal AI and point out that while the previous industrial revolutions — prompted by technological advancements — erased lots of jobs, they managed to create new ones (IT jobs, for example). So, new roles could flow from a legal AI revolution; these could be ‘Legal Technologist’, ‘Legal Hybrid’, ‘Legal Knowledge Engineer’ and ‘Enhanced Practitioner’.

These futuristic-sounding job titles are purely assumptions and guesses, nothing to bet on just yet. But, one thing we can be sure of is that AI is coming to the legal sector and it’s coming fast. We should not think anymore of ‘if’ but instead of ‘how’ AI will replace human lawyers.

Marco De Roni is a law graduate and paralegal in Amsterdam.