A recent survey published in The Economist estimates that 85 per cent of IT executives in banking have a “clear strategy” for adopting AI in the development of new products and services. Nevertheless, some business leaders continue to exercise caution. To promote the adoption of AI, Alex Luketa, partner at artificial intelligence (AI) data management specialist Xerini, tackles some key concerns regarding the implementation of AI and discusses the measures in place to overcome them.
The Information Commissioner’s Office (ICO) has stipulated that businesses must address the privacy risks associated with generative AI before adopting the technology. Stephen Almond, the ICO’s executive director of regulatory risk said, “We will be checking whether businesses have tackled privacy risks before introducing generative AI — and taking action where there is risk of harm to people through poor use of their data. There can be no excuse for ignoring risks to people’s rights and freedoms before rollout.”
For many in the financial services sector, AI privacy and the possibility of input data being used to train models are key concerns. No one wants their company’s private information to be available to competitors, and keeping clients’ financial data secure is critical. However, there are tools available that can improve the privacy of training data. For example, redaction tools that strip out personal or sensitive information like names and addresses, preventing anything from being sent to the large language model.
Where privacy is a more pressing concern, private models hosted on a finance organisation’s own infrastructure can help solve any issues. Working with a trusted AI partner who can consult on risks and mitigation measures can keep privacy concerns at bay.
Will it break the bank?
AI projects were historically less certain. It wasn’t always possible to guarantee a time frame or know for sure whether a model would work, which increased the risk of high upfront quotes or projects going over budget. Nowadays, there are fewer unknowns and a greater number of tools at a software engineer’s disposal. For example, Xerini developed Xefr, which works alongside existing processes and can be customised for more efficient integration.
AI has moved from the realm of specialist software development to become far more standard, making it a more competitive and affordable marketplace. The number of competing open-source services gives AI software consultants leeway when advising clients on the most cost-effective solution.
Will jobs be lost?
AI has the potential to enhance the effectiveness of many professionals by automating administrative tasks, allowing staff members to increase their overall productivity. For example, finance managers can use AI tools to analyse bank statements, tax documents and invoices, freeing them up to focus on client management and other tasks.
It’s natural for people to be concerned about how AI may impact their employment, so
when discussing the integration of AI, it’s crucial for software consultants and management teams to emphasise improvement rather than substitution. An effective AI implementation partner can help finance firms understand how to maximise their team’s potential with the help of AI tools.
Will it be biased?
The potential for bias removal ultimately hinges on the finance and banking data itself, a challenge amplified in a society with inherent biases. Software engineers can take steps to reduce the impact of biases by introducing guard rails, priming a system for bias reduction, or establishing a traceable answer source backlog. There is still work to be done, and bias in AI models will remain a topic of immense active research for some time yet.
Every business, no matter its size or current situation, can start its AI journey today. You can leverage readily available tools such as ChatGPT, integrate existing data and systems using customisable platforms like Xefr, or even create a custom private model. Partnering with an expert consultancy like Xerini can help you overcome implementation barriers and get the most out of AI.
About Author: Alex Luketa is a senior software architect with over 14 years of experience working in the financial sector, designing and building a variety of systems for clients including Morgan Stanley, Goldman Sachs and Credit Suisse. Alex is also experienced in the burgeoning field of artificial intelligence, and has developed many practical AI applications, including an innovative in-car app that uses machine learning to improve road safety.
