Although a scary concept for some, Artificial Intelligence (or AI) is well entrenched in today’s society.
From chatbots on websites, to facial recognition in airports and even self- driving vehicles, AI is changing the way we live. Generative AI (or Gen AI) is a subcategory of AI that uses algorithms to generate outputs like text, images or audio based on large quantities of data and, like all technology, has the potential to increase efficiency and save costs by minimizing human involvement.
Like many industries, the legal sector is beginning to embrace the use of Gen AI in the workplace.
Gen AI streamlines tasks such as document drafting, legal research and contract analysis, allowing legal professionals to focus on more complex matters. But as with any innovative technology, both AI and Gen AI come with its own set of risks.
Risk of data exposure
When using Gen AI tools, you need to be conscious of the information that you are inputting because once its out there, its impossible to get back. AI systems often require vast amounts of personal data, raising concerns about compliance with New Zealand’s privacy laws. Individuals and businesses alike must be alert to inadvertently exposing confidential data, particularly if the tool is not secure or is hosted on external servers.
Accuracy and reliability
Gen AI cannot understand or validate the accuracy of its output and can create nonsensical or false content known as “hallucination”. This was demonstrated by the use of fake case law by lawyers in America who failed to check that the cases actually existed, resulting in the Court dismissing their client’s case. Professionals are not absolved from legal responsibility or liability, or their professional obligations, if results produced by Gen AI tools are incorrect. AI may also fail to grasp the intricacies or legal requirements required in complex matters, leading to errors or oversight.
Bias and discrimination
The output from Gen AI tools is based on existing or artificially created input data. Being in its early days, certain groups may be underrepresented in the training data, causing AI to generate outputs that do not account for diverse perspectives or experiences. It also lacks the ability to understand societal nuances, cultural differences and evolving social norms, which can result in outputs that reflect outdated or narrow perspectives. From a legal context, AI outputs could result in bias against certain demographics in real-life situations such as in employment and healthcare, undermining fairness and equality from a human rights perspective.
Cybersecurity
The use of Gen AI tools can also create cybersecurity risks. Exploitation of vulnerabilities in the tools can give rise to data corruption or cyber-attacks on users. AI-generated phishing emails are
more convincing, bypassing traditional detection methods, potentially leading to data breaches or system downtime. And AI can generate very realistic audio and video, which can be used for impersonation, fraud or misinformation.
However, AI does also have some benefits when it comes to cybersecurity. AI can, itself, analyse data to identify patterns which are indicative of cyber threats and can adapt and reconfigure security protocols based on those threats.
Regulatory Response to AI
There have been varying regulatory approaches around the world to AI and the risks associated with it. A cabinet paper was released by the New Zealand Government in July 2024 which set out expectations for government agencies using AI, noting that AI can deliver better results for New Zealanders and agreeing that New Zealand “will take a proportionate and risk-based approach to AI regulation where needed” in preference to developing separate AI legislation. It recognised that there are existing laws to provide some protection, including those dealing with privacy, intellectual property and human rights.
However, it was agreed that overseas regulation policies should be observed and the OECD AI Principles promoted as being the direction for responsible AI use in New Zealand. AI promotes innovation and supports productivity and it is clear that AI has its place in the modern world. Having processes and procedures in place to minimize risk will maintain a balanced approach between utilising the benefits of AI and Gen AI while identifying the importance of the human role and accountability in the AI world.
Author: Anna is a Director at Bramwell Bate. Born and raised in Gisborne, Anna graduated from Victoria University with a Bachelor of Law and a Bachelor of Arts in Japanese. After practicing law in Gisborne and Wellington, Anna joined Bramwell Bate in 2015, working in corporate and commercial law, trusts, property, estates and succession law. Outside of work, Anna enjoys the Hawke’s Bay lifestyle with her family and friends. anna@bramwellbate.co.nz / 06 872 8210
