Why the finance industry is looking to agentic AI
Why the Finance Industry Is Embracing Agentic AI
Why the finance industry is looking – Imagine a scenario where an AI-powered assistant not only provides recommendations but also independently completes transactions. This is the essence of agentic AI—a type of software capable of executing intricate tasks autonomously, rather than simply responding to user requests. At the Money 20/20 Europe event in Amsterdam, industry leaders showcased the potential of this technology when Mastercard, ING, and Worldline announced the completion of Europe’s inaugural live, end-to-end agentic payment system. The demonstration highlighted a seamless process: a consumer instructed an AI agent to search for concert tickets in a specific location on a set date, within a predetermined price range. The AI identified viable options, executed the payment with human oversight, and finalized the transaction. This milestone underscores how agentic AI is transforming the financial landscape.
The Evolution of AI in Finance
For years, agentic AI has been a buzzword in financial technology circles, but its practical implementation is now gaining traction. Scarlett Sieber, chief strategy and growth officer at the conference, noted that AI adoption has moved beyond experimental ventures and is now being integrated across major institutions. The University of Cambridge’s 2026 report, which surveyed over 600 firms and regulators globally, projects that the deployment of AI agents in finance will increase from 24% to 81% by 2030. While this growth is promising, the report also warned that the pace of innovation may outstrip existing oversight mechanisms, leaving gaps in technical readiness and regulatory frameworks.
Agentic AI is not just a theoretical concept—it is already reshaping operations. Take eToro, the Israeli-based fintech company, which has made significant strides in leveraging AI for investment decisions. Its CEO, Yoni Assia, shared how the platform’s AI assistant has evolved from offering portfolio insights to acting on behalf of users under predefined limits. A standout example is the POTU$ app, which analyzes Donald Trump’s social media activity and breaking news to detect market-moving events. When the U.S. president shares content that could impact financial markets, the AI swiftly places trades in user accounts, often within seconds of the post. This capability demonstrates how agentic AI can bridge the gap between real-time data and immediate action.
Automation and Efficiency in Financial Services
Meanwhile, Klarna, the Swedish fintech firm behind the “buy now, pay later” payment options seen on numerous e-commerce platforms, has also embraced AI to streamline operations. Last month, the company launched a shopping search application within ChatGPT, expanding its digital footprint. In 2024, Klarna partnered with OpenAI to develop an AI assistant that handles customer service tasks, claiming it replaced the workload of 700 full-time human agents. The CEO, Sebastian Siemiatkowski, emphasized that AI enables the organization to achieve more with fewer resources, as its workforce has decreased from 6,000 to fewer than 3,000 in recent years, while revenue per employee has risen. However, Siemiatkowski acknowledged the trade-offs, noting that cost-cutting efforts had initially led to a decline in service quality. To address this, Klarna has begun rehiring human agents and investing in their role as a critical component of the customer experience.
“AI is useless without humans steering it,” Siemiatkowski said, reflecting on the balance between automation and human oversight.
Traditional financial institutions are also adapting to this shift. ABN AMRO, the third-largest Dutch bank, has seen a dramatic reduction in physical branches, from 500 in 2010 to just 26 today. CEO Marguerite Bérard explained that AI is a cornerstone of the bank’s broader digital transformation, with 85% of its staff now using AI tools daily. The bank’s AI bot “Ana” facilitates millions of customer interactions, while “Lenny” automates credit request processes. These examples illustrate how agentic AI is not only a tool for innovation but also a driver of efficiency in legacy systems.
Challenges and Uncertainties
Despite its promise, the widespread adoption of agentic AI has sparked concerns. Gartner, a leading research firm, predicted that over 40% of agentic AI projects could be canceled by the end of 2027, citing issues such as rising costs, ambiguous business value, and insufficient risk management protocols. A recent collaboration between Accenture and Wharton Business School highlighted the need for clear governance structures, emphasizing that leaders must decide which decisions to delegate to AI and where human judgment remains essential. “The integration of agentic AI requires a deliberate approach to ensure accountability and trust,” the report stated.
Industry experts stress that while AI is automating tasks, it cannot fully replace the human element. Bérard of ABN AMRO reiterated this point, explaining that AI’s effectiveness depends on the quality of the processes it supports. “If you deploy AI on a flawed system, you’re still dealing with inefficiencies,” she said. This highlights a key challenge: ensuring that AI enhancements are built on solid foundations rather than rushed implementations.
The Future of Work in Finance
The rise of agentic AI is also redefining job roles. Siemiatkowski of Klarna warned that automation could lead to significant workforce changes, particularly in customer-facing positions. He noted that sales roles and legal functions may be most vulnerable, though some areas could see short-term disruptions. “There might be negative impacts in specific job categories,” he admitted, though he did not specify the percentage of roles replaced by AI. This raises questions about the long-term implications for employment in the sector.
Assia of eToro, however, emphasized that AI does not eliminate the need for human input. “Our AI assistant is a tool, not a replacement,” he said. The platform’s growth in AI capabilities—from basic insights to autonomous actions—has been exponential. In six months, AI usage across the company increased tenfold, and 95% of new code was developed by AI systems, compared to none two years prior. This rapid integration reflects a broader trend where financial institutions are leveraging AI to reduce costs and enhance scalability.
Industry Collaboration and Innovation
Traditionally, fintech startups and conventional banks were viewed as competitors. However, the Money 20/20 Europe event revealed a shift toward collaboration. Many institutions are now pooling resources to develop agentic AI solutions that can compete with or complement each other. This partnership is crucial as the technology matures, with firms working to address challenges like regulatory compliance and technical limitations. As Sieber noted, the focus has moved from hype to tangible results, with real-world applications now driving the conversation.
The integration of agentic AI into everyday operations is expected to redefine customer expectations. Whether it’s a bank’s AI-powered credit assessment or a fintech app’s ability to act on market trends, the technology is enabling faster, more personalized services. Yet, its success hinges on the ability to balance innovation with oversight. As the University of Cambridge report highlighted, the financial industry must ensure that AI’s capabilities are matched by robust frameworks to manage risks and maintain transparency. This delicate equilibrium will determine how widely agentic AI is embraced and its long-term impact on finance and society.
With agentic AI becoming a central focus, the question remains: how will this evolution shape the future of financial services? The examples from Mastercard, eToro, and Klarna suggest that the technology is already making waves, but the journey is far from complete. As institutions continue to refine their AI strategies, the interplay between automation and human expertise will define the next phase of innovation. For now, the financial sector is not just observing the potential of agentic AI—it is actively shaping its role in a rapidly changing world.
