Analysing the future of AI in law
There is much optimism (and some trepidation) about the possible impact that artificial intelligence (AI) could have on the practice of law.
On the one hand, there is the idea that AI could provide significant efficiencies to lawyers, make the law and courts more accessible to the layperson, and reduce costs and delays. On the other hand, there is a fear that lawyers will be rendered redundant and replaced by machines. The reality is that while both could eventually prove to be true, at least to some degree, there are a number of hurdles still to overcome before AI can achieve its potential in the legal arena. Furthermore, even if that occurs, the early indications are that there will be some aspects of legal practice that will always be better suited to being carried out by a human.
To date, the impact of AI on the legal profession has been limited to areas such as discovery, research and the review of contracts. These are all areas involving relatively large datasets and a feedback loop that can be used to train the algorithms underpinning AI. The discovery process, for example, which often involves a person sifting through many thousands of documents and making an assessment as to their relevance to legal proceedings, is particularly well suited to benefiting from the application AI. Indeed, discovery software is already in use that relies on a sample of reviewed documents to predict to a high degree of accuracy whether unreviewed documents are also going to be relevant.
There is also some optimism that AI may be able to play a role in better quantifying litigation risk. Where a large enough dataset exists, it may be soon be possible to predict the chances that a particular case succeeds in court by assessing the key facts of that case against reported decisions of a substantially similar nature.
The promise of AI then (and it is, to date, something of an unrealized promise) is that it could be used to more efficiently achieve tasks that are time intensive for humans and that it can more precisely predict future outcomes.
If anything, these would seem to be tools to better equip lawyers to do the things that they do uniquely well (such as thinking and advising) as well as ensuring that they target those efforts towards cases and clients with the best prospects of achieving the desired outcome.
Perhaps most significant for the future of AI and the law though, is that law governs interactions between humans, who are emotionally complex and endlessly unpredictable. Much of the law (and particularly litigation) operates in the grey areas, seeking certainty where the facts fall somewhere between black and white.
As things stand, humans remain uniquely equipped to recognize, understand and respond to the humanity (motivations, desires, aspirations, fears) in others. These are factors which mean lawyers are uniquely placed to serve and represent their clients. There is undoubtedly almost unlimited potential to the way that AI can be used as a tool to assist the legal profession, but the idea that AI could replace humans working in the legal field altogether seems, to say the least, extremely premature.