By Alex Dawson
Artificial intelligence, the new corporate buzzword. Whilst it appears to be ‘the Goose that Laid the Golden Eggs’, its abilities need to be taken with a pinch of salt. Whilst this article does not advocate the killing of this Goose (AI is an essential competitive edge to all firms), it does seek to dig deeper into its future use in the legal and financial service sectors.
The Use of AI in the Legal Sector
The most common uses of AI within the legal sector are document review, document management, and document and precedent discovery. In essence, it automates more mundane and repetitive tasks previously completed by associates. However, the use of AI has begun to develop into 24/7 chatbots able to answer frequently asked legal questions and even analyse cases to predict their chance of success (with one provider claiming 90% accuracy).
The benefits for law firms are immense. In an economic climate demanding efficiency and reduced legal costs, automation saves time and human capital. This means greater resources can be allocated to solving more complex problems. This further efficiency translates into law firms having a greater client capacity, resulting in increased net profit.
The Limits to AI’s Further Use in the Legal Sector
The logical next development is for lawyers to be able to input a legal situation into an AI model which will generate the perfect solution. However, this is fraught with barriers.
Primarily, confidentiality is key. The Solicitors Regulatory Authority advises the disclosure of the use of AI to clients as the use of AI facilitates the risk of data breaches. In a heavily saturated legal market with many firms having similar USPs, trust is key. If a data breach occurs, it is foreseeable a firm would lose many of its clients due to the loss of trust. This is a substantial risk, particularly in an economic downturn where law firms likely have reduced reserve capital to subsidise this loss of revenue.
Secondly, AI models are prone to ‘hallucinations’. Whilst it may be humorous to read about a lawyer who used ChatGPT and referenced non-existent precedents (see the New York lawyer on trial for such an occurrence), it is an inaccuracy that law firms cannot afford. To damage their reputation is to lose clientele.
Therefore, until AI becomes more trustworthy and accurate, its further use will be restricted.
The Use of AI in the Financial Services Sector
Financial service firms have similarly welcomed the use of AI, with a reported 90% of firms incorporating AI. Like their legal counterparts’ mundane tasks have been automated, such as to improve online banking transactions and anti-money laundering schemes (notably Wells Fargo’s ethos is “Nobody gets a dopamine hit from banking a deposit”). However, it has also taken on a much larger role. BlackRock, JPMorgan Chase, and Morgan Stanley use AI to pick stocks and manage their portfolios. AI has been trained to analyse data, historic market trends, and market sentiment to trade stocks quicker than its human counterparts.
In 2023 Investopedia put man against machine for three months, with machine achieving double the returns of man. This speed and accuracy of investment will increase as these models are given larger data sets. This ultimately translates into greater profitability, and lower risk for these firms, which in turn attracted customers.
The Limits of AI’s Use in the Financial Services Sector
Whilst it is difficult to establish a comprehensive assessment due to AI’s rapidly evolving use, the greatest present risk is cyber security. A prompt injection attack allows for the alteration of a prompt given to AI models, without the knowledge of its user. Consequently, the model will output incorrect responses, even when specifically programmed against them. This would result in substantial losses. Additionally, this risk is not something that can be economically insured against. Therefore, from a risk management perspective, it is not a sustainable approach. Consequently, until these vulnerabilities can be resolved (or economically insured against), further development will be slow.
Conclusion
AI has significant barriers in the legal and financial service sectors. There are issues of confidentiality, trust, hallucinations, cyber-security threats and the lack of insurability. Whilst AI is certainly the ‘next big thing’, it may not be such an immediate revolution.
Comentarios