Posted by Andrew Davies, managing director of Legal Futures Associate SpeechWrite 
In my blog last month , I looked at how developing forms of artificial intelligence (AI) are increasingly being used in everyday life, as well as helping business processes become more streamlined. But how, specifically, can AI help the legal sector?
The current state of AI
While we’re years away from truly autonomous forms of AI (the type of ‘strong’ AI Sofia the robot  represents an early step towards), weaker or ‘narrow’ AI is already enhancing our lives.
Google is probably the most advanced and ubiquitous form of AI we’re familiar with – using natural language processes to interpret users’ search queries (spelling mistakes and all) in order to retrieve the most relevant results.
Here are just a few of the ways intelligent technologies are helping to transform how law firms manage their practice.
Accurately logging billable hours is the goal of most law firms – and, nowadays, intelligent time recording tools can do this. Plus, at the end of each month, they can automatically generate invoices, helping to reduce admin hours and streamline law firm processes.
Digital dictation software
Sophisticated digital dictation  not only responds to voice input with speed and accuracy, but it can also provide advanced customisation options, transcription ability and seamless integration across other devices.
For example, this type of software is capable of pulling in information from CRM databases and also identifying potential upselling opportunities.
What’s more, it is becoming increasingly intuitive, with voice recognition devices adapting to users’ individual voices. It is also starting to incorporate the type of command elements normally associated with virtual personal assistants, such as Apple’s Siri – making it possible to create documents, send email, search online and control a PC simply by using voice commands .
Predictive coding and e-disclosure
This is one of the most advanced forms of AI utilised by law firms, using a teachable search algorithm which learns to rank the relevance of documents.
Technology-assisted review software is initially taught by a lawyer to assess the relevance of any particular document. Then it’s let loose to hunt through huge volumes of electronic files for anything of relevance, at the fraction of the time and cost it would take a paralegal team to do the same job.
With around 50 million words in the statute books (according to the National Archives ), the ability to represent legislation as data has transformed the way lawyers can hone in on the most relevant sections.
Online information repositories like LexisNexis are constantly evolving their search algorithms to enable lawyers to find the most relevant material, while tools like Lex Machina can even help lawyers build a case strategy based around outcomes in similar cases.
DIY law and chatbots
Imagine being able to assist hundreds (or even thousands) of clients a day without ever having to speak to them. This is what DIY document assembly software and chatbot tools can help firms achieve.
They allow people to create their own legal documents or access basic legal advice without a lawyer. Like Parker – a chatbot ‘employed’ by Norton Rose Fulbright in Australia . This knowledgeable bot is helping to free up fee-earners’ time by answering simple questions from clients.
The future of legal AI
A study last year  by LawGeex, a legal tech start-up, challenged 20 lawyers to test their smarts against LawGeex’s own AI-powered algorithm.
The task was to review risks contained in five NDAs. The AI matched the top-performing lawyer for accuracy (both achieving 94%; the average overall was 85%). And when it came to speed, the AI took just 26 seconds to review all five documents, compared to an average of 92 minutes by the lawyers.
Oh, and the AI didn’t drink any coffee either, compared to the 12 cups consumed by the lawyers!
Of course, there are things AI can’t do – like meet with clients or appear in court. However, the potential of utilising AI to make gains is significant, especially when it comes to data-crunching tasks.
So law firms need to think holistically, and start asking themselves how, and where, AI can fit into their practice, and what it can contribute to the delivery of their service.