Role of AI in reshaping financial service

By Karen Jones, Founder and CEO of Citywealth, a media & events platform for the global wealth management industry

£billions to be spent just getting ready for AI

With the proposal for the EU Artificial Intelligence Act now available and slated to come into force in less than two years, it might be time for finance professionals to start preparing.  Like the run up to GDPR, there is a period of time to get ready for compliance. That is a substantial piece of work for all businesses for a couple of reasons. Mainly, everyone uses AI, and there is AI at scale. This is because wealth managers and other types of financial institutions, have hundreds, if not thousands, of AI deployments and they often are unable to point to a list; it requires some work for them to answer that question and they are often surprised at just how many deployments they actually have. The future holds a vast compliance job. The European Commission estimates €7bn in annual compliance costs, so this is not a trivial addition or obligation.

Legal confidentiality

So, when it comes to administrative work and data management jobs, AI seems already to be playing a substantial role in making an efficient dent in the ‘grunt work’. But is there any truth to the news headlines claiming lawyers and other professionals will be edged out of their jobs by AI? Some platforms offer ‘sentiment analysis’ which assesses the type of emotion expressed in a document. Obviously, there must be proper human oversight, but AI does open up these sorts of concepts. However, currently law firms can use AI to trawl through large amounts of text and data from public sources, but not to generate text using information about clients or counterparties. It seems legal confidentiality will reign supreme over AI. It seems there are serious data and intellectual property issues to navigate.

Investment management

AI portfolios base their decisions on advanced algorithms and data analysis. They can process vast amounts of data from various sources, identifying patterns and correlations. The predictive analysis models aim to predict expected returns on different asset classes to generate investment ideas or arbitrage opportunities. Risk management in portfolio management activities is another area in which AI can help by showing different outcomes or scenarios.

Despite the gains though, just like in the legal sector it is important to be aware of limitations or the risks it presents in portfolio management. First, it is essential that data fed into AI is of high quality. Otherwise, the results lack reliability and at worst could start a systemic ‘run on the bank’.

One area that is positive is sentiment analysis which is the process of analysing digital text to determine if the emotional tone of the message is positive, negative, or neutral. It can help companies understand more quickly customer support chat transcripts, social media comments, and reviews.

Hallucinations

Some professional services firms are attempting to use AI robots to provide curated legal advice. The results so far have been patchy as large language-based AI robots tend to ‘hallucinate’ – in other words they often make up plausible sounding, but wholly incorrect legal advice which has no basis in reality. This is why advice issued by professional services firms that is derived from AI robots has to be curated by a human.

Some think AI enterprise licences will be how we use it through third parties like Microsoft who essentially run it through their software and will absorb the risks for us. Bloomberg have launched BloombergGPT which is trained on financial data to support tasks in the finance sector. They will be using sentiment analysis, named entity recognition, news classification and question answering. It will also use its data with AI to help existing customers more.

Time to be more human

AI is currently more focused on document-heavy transactions. However, AI and machine learning are now also being used to produce standard form agreements, negotiate basic contracts, and produce legal analysis. Whilst it does not come without challenges, AI has an important role in the future of the wealth management industry alongside the trusted advisor relationship, to whom clients still look for comfort on advice.

Research by Accenture after surveying 500 financial advisors across the U.S. and Canada, found 83% ‘believe AI will have a direct measurable and consistent impact on the client-advisor relationship in the next 18 months.’ This can translate into better informed wealth managers that can pass key insights to their clients while retaining the qualitative factors financial advisors provide. By taking on repetitive tasks, AI can help to significantly reduce costs and enhance the client-advisor relationship, giving advisors back precious time to focus on the human interaction and close connection to their clients.