October 1, 2026 – The financial services sector is abuzz with the potential of agentic artificial intelligence (AI), a sophisticated form of AI designed to autonomously perform tasks, learn, and make decisions. However, the critical question facing institutions today is the extent to which this widespread enthusiasm is translating into tangible, impactful enterprise-wide deployments. A recent expert roundtable, convened by GlobalData, delved into this very dichotomy, dissecting the opportunities and significant challenges hindering the widespread adoption of agentic AI within banking and financial services.
The discussion, aptly titled “Agentic AI in Banking & Financial Services: Beyond the Hype,” brought together a distinguished panel of industry leaders: Jeff Veis, Chief Marketing Officer at Impetus Technologies; Deepak Khosla, Chief Growth Officer & Head of AI Business at Impetus Technologies; and Stephen Walker, Retail Banking Analyst at GlobalData. Their collective insights illuminated a landscape where experimentation is rife, but scaled production remains an elusive goal for many.
The Chasm Between Experimentation and Enterprise Impact
While financial institutions are actively exploring agentic AI’s capabilities, a significant gap persists between pilot projects and full-scale integration into core business operations. The roundtable participants identified several critical barriers contributing to this lag:
- Data Readiness: A fundamental prerequisite for any AI system, particularly agentic AI, is the availability of high-quality, well-structured, and accessible data. Many financial institutions grapple with legacy data systems, data silos, and inconsistent data governance, creating an environment where AI agents struggle to access and process the information they need effectively.
- Fragmented Enterprise Context: Agentic AI thrives on a comprehensive understanding of the business environment. The complexity of financial services, with its intricate workflows, diverse product offerings, and nuanced regulatory frameworks, often results in a fragmented understanding of the enterprise context for AI systems. Without this holistic view, agents can operate with incomplete information, leading to suboptimal or even erroneous decisions.
- Governance Gaps: The autonomous nature of agentic AI raises significant governance concerns. Establishing clear lines of accountability, robust oversight mechanisms, and ethical guidelines for AI-driven decisions is paramount. Existing governance structures in many financial institutions are not yet equipped to handle the complexities introduced by agentic AI.
- Safety Risks: The potential for unintended consequences and errors in highly regulated and sensitive financial operations is a primary concern. Ensuring the safety and reliability of agentic AI agents, especially in critical areas like credit, fraud detection, and customer interactions, requires rigorous testing, validation, and continuous monitoring.
- Regulatory Expectations: Regulators worldwide are closely scrutinizing the deployment of AI in financial services. The lack of clear, harmonized regulatory guidance on agentic AI, coupled with the inherent opacity of some AI models, creates uncertainty and hesitation among financial institutions. Meeting evolving compliance demands while innovating with AI presents a formidable challenge.
Redefining the Role of AI Agents in Finance
Deepak Khosla articulated a crucial distinction about the sophistication required for agentic AI in the financial sector. He stated, "An agent in banking is not just summarizing a document. It could influence and impact credit, fraud, payments, customer treatment, reporting, or advice. The bar for production is therefore much higher in the banking and financial services sector." This statement underscores that agentic AI in finance is not merely an efficiency tool; it is a potential driver of strategic decisions with far-reaching consequences.
The implication is that financial institutions cannot afford to treat agentic AI deployments with the same approach as less critical applications. The stakes are considerably higher, demanding a more robust, secure, and intelligent implementation.
The Imperative of Enterprise Context Engineering
Khosla further emphasized that successful agentic AI adoption hinges on establishing a strong, AI-ready data foundation. This means more than just having data; it involves grounding AI agents and systems in the proprietary business processes, operational realities, historical interactions, and crucially, the governance policies that define a financial institution.
"Our approach starts with the belief that agentic AI success depends on the quality of enterprise context available to agents," Khosla explained. "If that context is fragmented, stale or poorly governed, they will fail in production." This highlights a fundamental truth: AI agents are only as effective as the information and understanding they are given. In the complex world of finance, this context needs to be meticulously curated and continuously updated.
The roundtable concluded with a clear consensus: financial institutions must prioritize the engineering of enterprise context. This involves building trusted semantic layers, ontologies, and knowledge graphs. These structured frameworks provide AI agents with a reliable and consistent understanding of the business domain, enabling them to reason effectively over complex data and scenarios. By investing in context engineering, financial firms can unlock strategic, high-impact use cases where the return on investment is clearly measurable and the associated risks can be effectively managed.
A Timeline of Evolving AI Adoption in Finance
The journey towards agentic AI in financial services can be traced through several key phases:
- Early 2020s: The Dawn of Generative AI and Initial Exploration: The widespread availability of powerful Large Language Models (LLMs) sparked initial interest in AI’s potential for tasks like content generation, summarization, and basic customer service automation. Many institutions began experimenting with these foundational AI technologies.
- Mid-2020s: The Rise of Agentic Concepts: As the limitations of standalone LLMs became apparent, the focus shifted towards more autonomous AI agents capable of performing multi-step tasks and interacting with systems. This period saw increased research and development in agentic AI frameworks and tools.
- Late 2020s (Present): The "Beyond the Hype" Phase: As evidenced by the GlobalData roundtable, the current phase is characterized by a critical assessment of agentic AI’s real-world applicability and scalability in finance. Institutions are moving beyond theoretical possibilities to confront the practical challenges of deployment. The emphasis is shifting from "can we" to "how do we effectively and safely implement."
- Future Outlook (2027 onwards): Maturation and Strategic Integration: As the industry addresses the current barriers, the future is expected to see more mature and strategically integrated agentic AI solutions. This will likely involve robust governance frameworks, standardized data practices, and a deeper understanding of how AI agents can augment human capabilities in critical financial operations.
Supporting Data: The AI Investment Landscape
While the exact figures for agentic AI investment are nascent, broader AI investment in financial services offers indicative trends. According to a recent report by IDC, worldwide spending on AI systems in the banking industry is projected to reach $40.3 billion by 2027, growing at a CAGR of 19.7% from 2023. While this encompasses all AI applications, the increasing sophistication of these investments points towards a growing allocation for more advanced capabilities like agentic AI.
Furthermore, a survey by Deloitte in late 2025 revealed that 70% of financial services executives reported having an AI strategy in place, with a significant portion indicating plans to increase their AI investments over the next two years. This widespread strategic commitment underscores the underlying momentum driving the exploration of agentic AI, even with the current deployment challenges.
Reactions from Related Parties and Industry Analysts
While specific company reactions to the GlobalData roundtable were not detailed in the initial announcement, the insights provided by Stephen Walker, Retail Banking Analyst at GlobalData, offer a broader perspective. His role as an analyst suggests that such discussions are crucial for informing the wider market about the practical realities of emerging technologies.
"The conversations at events like these are vital," a hypothetical industry observer might comment. "They help to temper the unrealistic expectations often fueled by vendor hype and guide institutions towards pragmatic strategies. Focusing on foundational elements like data and context is a common refrain across various technology adoption cycles, but it’s particularly critical for autonomous systems like agentic AI."
Moreover, regulatory bodies are also keenly observing the advancements. The Financial Stability Board (FSB) and various national central banks have been actively researching the systemic risks associated with AI, including autonomous systems. Their ongoing work in developing frameworks for AI governance and risk management will significantly influence the pace and nature of agentic AI deployment in the coming years.
Broader Impact and Implications for the Financial Sector
The successful navigation of the challenges surrounding agentic AI adoption will have profound implications for the financial services industry:
- Enhanced Efficiency and Productivity: Once deployed effectively, agentic AI can automate complex processes, freeing up human capital for higher-value tasks. This could lead to significant cost savings and improved operational efficiency.
- Improved Customer Experience: Agentic AI can power highly personalized customer interactions, proactive advice, and seamless service delivery across multiple channels, potentially leading to increased customer satisfaction and loyalty.
- Smarter Risk Management: In areas like fraud detection and credit assessment, agentic AI, grounded in robust context, can identify patterns and anomalies with greater speed and accuracy than traditional methods, leading to more effective risk mitigation.
- Innovation in Product Development: The ability of agentic AI to analyze vast datasets and identify market trends could accelerate the development of innovative financial products and services tailored to evolving customer needs.
- Reshaping the Workforce: The increased adoption of agentic AI will undoubtedly lead to a shift in the skills required within financial institutions. There will be a growing demand for AI specialists, data scientists, and professionals capable of managing and overseeing AI systems.
The roundtable’s emphasis on "context engineering" suggests a path forward. By investing in the foundational elements that enable AI agents to understand and operate within the intricate web of financial services, institutions can move beyond the hype and unlock the true, transformative potential of agentic AI. The journey is complex, but the rewards—for both financial institutions and their customers—are substantial.
