The integration of AI in various industries has been a topic of immense interest and discussion in recent years, with the banking sector being no exception. AI offers the promise of streamlining operations, providing personalized services, and enhancing overall efficiency.
However, as discussed during Money Motion 2024 by industry experts, the technology also brings forth a host of challenges that must be carefully navigated for successful implementation.
Automation and personalization
AI has the potential to revolutionize banking by automating routine tasks, thereby freeing up human resources to focus on more complex and strategic initiatives. Drazenko Kopljar, COO at PBZ, highlighted the potential benefits but also emphasized the significant challenges ahead. He pointed out that internal and external resistance, along with regulatory concerns, are primary hurdles that must be addressed.
Hajdi Ćenan, CEO of airt, a Croatian deep tech startup whose deep learning platform helps companies predict their end customers’ behavior, noted that recent advancements in computing power and data availability have accelerated AI implementation. The emergence of Generative AI (GenAI) and AI agents signifies a shift towards more sophisticated and capable AI systems.
“Now we also finally have computing power that could also analyze and crunch all that data. What is happening now with ChatGPT and the LLMs is the case of progression. Because if you look at AI developments, what we were talking about before was about machine learning and deep learning. GenAI is kind of a subset of just another approach, another technique of analyzing that data,” Ćenan said during the AI and data driven banking panel at MoMo24.
These AI agents, powered by machine learning and deep learning techniques, hold the promise of handling transactions within banking applications, she points out.
“I think that ChatGPT or AI agents will be able to handle transactions as well, because you will create agents that will use deep learning or machine learning techniques. Therefore, you will have multi-agent AI systems where you’re having specialized agents who have to communicate with each other and when that happens, you’ll have much more control over what a single agent can do,” Ćenan explains.
Ethical considerations and data quality
While the potential benefits of AI in banking are vast, it is crucial to address the challenges that come with its implementation. Ethical considerations surrounding AI-driven decision-making processes, data privacy, and bias mitigation are paramount. Nenad Crnčec, founder at Architech, emphasized the need for caution, stating that while AI can automate processes, critical decision-making roles still require human oversight.
“It will not be a decision-making role, but we will need some agents to be able to automate certain processes. But again, probably it will take some time before we do stuff such as automated risk analysis etc,” Crnčec added.
Access to high-quality data is another critical challenge highlighted during the discussions. AI systems rely heavily on data inputs for training and decision-making, and ensuring the accuracy, relevance, and security of this data is essential.
Without robust data governance frameworks and data quality assurance measures, AI systems may produce inaccurate or biased results, leading to potential risks and regulatory concerns.
Promising future, but under a lens of scrutiny
Despite the challenges, the consensus among experts is that the future of AI in banking is promising, provided that stakeholders collaborate effectively and prioritize ethical considerations and data quality. Collaboration between AI agents and human experts can lead to more efficient decision-making processes, enhanced customer experiences, and improved risk management strategies.
Continuous innovation and investment in AI research and development will also play a crucial role in shaping the future landscape of banking. And as experts emphasized, AI technologies evolve, incorporating explainable AI techniques and transparency measures will enhance trust and acceptance among customers and regulators.
