The AI FP&A Committee convened for its 26th meeting, marking its seventh anniversary. The session focused on AI agents in FP&A and featured an insightful case study titled "AI Agents in FP&A" by Peter Galik, Global Head of Innovation, Analytics and Automation at Takeda.

Peter Galik’s presentation explored the emerging role of AI agents in FP&A, focusing on practical applications and implementation challenges.

Peter outlined a structured framework for understanding AI agents, categorising technologies into:

1. Enterprise Platforms – Foundational ERP and CRM systems.
2. Data Infrastructure – Data warehouses, data lakes, and analytical models.
3. General IT Infrastructure – Cloud computing, mobility, and network solutions.
4. Process Automation – RPA and workflow automation solutions.
5. Emerging Technologies – AI, machine learning, IoT, blockchain, and generative AI.

He emphasised that AI agents, particularly generative AI, have introduced a paradigm shift by enabling systems to ‘understand’ and automate complex workflows, reducing manual intervention in FP&A.

He highlighted two primary applications of AI agents: AI-Powered Assistants that automate routine queries, transactional tasks, and self-service processes, and Back-Office AI Agents that automate decision-making in tasks such as invoice processing, financial analysis, and predictive modelling.

The discussion addressed key challenges in AI adoption for FP&A. A major concern is data quality — AI is only as good as the information it learns from, and unreliable data leads to unreliable insights. While generative AI is improving, it still has a tendency to produce misleading results, making validation crucial. The role of finance professionals is also evolving, requiring them to understand and oversee AI-driven processes rather than relying on automation blindly. Another challenge is AI’s struggle with numerical accuracy; despite advancements, large language models still have difficulty handling complex financial data. 

The meeting concluded with an engaging discussion on the future of AI in FP&A. Key insights included:

• AI agents are already enhancing FP&A efficiency, but they need to be better integrated into enterprise systems for organisations to realise their full potential.
• Adapting to AI isn’t just about technology — it is about culture. Finance teams need to embrace continuous learning and be open to new ways of working.
• The Committee will keep tracking AI developments, fostering knowledge-sharing, and supporting best practices across the industry.

If you would like to learn more about AI Agents in FP&A, you are welcome to read the article written by Peter Galik. To watch the videopresentation, please follow this link.