In a recent interview, Stuart Tait, Chief Technology Officer for Tax and Legal at KPMG UK, likened the introduction of generative AI solutions in finance to “moving from typewriters to word processors.” This striking analogy effectively captures the transformative potential of these technologies. Generative AI is already delivering tangible value across a wide range of financial tasks. Moving forward, it looks poised to become even more prominent as its adoption continues to expand within this important sector.
The excitement surrounding generative AI has reached considerable levels. Many believe these tools will significantly enhance operational efficiency within the financial sector, improve customer experiences, and streamline regulatory compliance. While this is certainly possible, there remain potential obstacles that could hinder progress. One of the most critical challenges is maintaining a consistent supply of clean, relevant, and accurate data, especially as businesses increasingly depend on AI for decision-making purposes.
DATA IS THE LIFEBLOOD OF AI
It’s fair to say that in the midst of all the excitement, the critical role of data in ensuring reliable AI systems has been somewhat overlooked. AI is often depicted as a near-magical, sentient technology that functions independently, without the need for input or assistance. However, this portrayal is far from reality. At the core of every AI system lies data, aptly described as the ‘lifeblood’ of these technologies. Consequently, if the data is imprecise, incorrect, or irrelevant, the AI systems built upon it will inevitably inherit these flaws.
In the finance sector, this crucial consideration cannot be ignored. Companies looking to develop AI systems must ensure they are doing so with the high-quality data that these systems require. Moreover, to truly unlock the full potential of AI in finance, businesses need systems that provide easy access to relevant data, ensuring that the data is correctly formatted for integration into AI systems. Without these systems in place, AI systems will struggle to deliver the long-term, transformative impact that many commentators seem to already assume is inevitable.
DELIVERING HIGH-VALUE INSIGHTS
To this end, I would find myself agreeing with Rohit Sehgal, Founder and CEO of Vincilium, who recently espoused that “AI needs data more than data needs AI”. AI’s true value lies in elevating the analysis and insights derived from high-quality data. Ultimately, if the data provided to these systems is lacking, difficult to utilise, interpret, or access, then poor outputs should be expected. In the realm of finance, this renders systems unusable and exacerbates the very problems they were designed to solve.
In an era of stringent regulations and heightened compliance demands, such an outcome could be particularly damaging. Inaccurate predictions, biased outcomes, and flawed decision-making could place financial institutions in serious jeopardy, potentially harming customer relationships and leading to costly fines. To prevent this, companies must ensure that AI systems are trained exclusively on high-quality, diverse, and comprehensive data sets. Unfortunately, sourcing such data in today’s complex environment can be a challenge.
DATA SOURCING CHALLENGES
Financial institutions often face significant hurdles in accessing valuable data due to legacy systems. These outdated platforms, which still store vast amounts of critical information in a fragmented, siloed form, can be difficult to integrate with modern data and AI systems, creating a barrier to effective data utilisation. Additionally, data silos pose a major challenge, as they fragment information across different departments or systems, leading to inconsistent or incomplete datasets. This fragmentation can significantly hinder AI development efforts.
Moreover, the struggle for clean data is an ongoing issue. Financial data is often messy, unstructured, or outdated, requiring extensive cleaning, organising, and structuring before it can be effectively used. This process is time-consuming and complex, but essential for ensuring that AI systems can deliver precise and valuable insights, rather than being undermined by poor data quality. Thankfully, data management systems that can access data flows regardless of format, system or silo can help to ease this concern.
FINDING THE RIGHT APPROACH
To build AI solutions that financial institutions can genuinely trust, the first and foremost requirement is to source data that is reliable and trustworthy and in an easily accessible format. This step is indispensable in the development process and warrants greater recognition from companies dedicated to achieving this objective. The capability to consolidate and aggregate large volumes of data is crucial in this context, as it enables AI systems to efficiently and effectively discover, analyse, and extract actionable insights.
With these solutions in place, financial institutions can confidently accelerate their broader AI adoption plans and start realising the transformative benefits that many anticipate this technology will deliver. By implementing robust data management techniques, companies can ensure the provision of accurate and up-to-date data, which is essential for systems to automate routine tasks previously performed by humans. This not only saves countless hours but also enables more precise predictions and enhances decision-making.
About the Author
Karthik Jagannathan is the Head of Payments Advisory at Intix, a global leader in transaction data management. With a proven track record of leading business and technology teams at major banks and technology firms, Karthik specializes in building innovative payment solutions. His deep expertise in payments, combined with extensive experience in managing multinational projects, enables Intix to continue leveraging industry-leading data management techniques, providing companies with clear, comprehensive, and actionable insights at their fingertips. Bringing over 20 years of experience in the financial services sector, Karthik is a well-recognized expert on instant payments and open banking in the industry. He possesses deep insights into compliance, data processing, and how organizations can harness technological innovation to drive meaningful changes in payments that directly benefit end-users. At Intix, he focuses on helping clients navigate global regulatory changes in an era of heightened compliance.
Sign up for the free insideAI News newsletter.
Join us on Twitter: https://twitter.com/InsideBigData1
Join us on LinkedIn: https://www.linkedin.com/company/insideainews/
Join us on Facebook: https://www.facebook.com/insideAINEWSNOW
