Artificial intelligence (AI) is spreading like wildfire in financial services, particularly for financial planning and analysis (FP&A) use cases. Adoption is high, with a recent NVIDIA survey reporting that 91 percent of financial services companies are actively evaluating or using AI to automate tasks and improve operational efficiency.
It’s no surprise that AI is so popular when it comes to FP&A workflows. AI tools can free teams from the drudgery of repetitive tasks and boost forecasting and analysis, allowing finance staff to focus more on high-value tasks and strategic decision-making.
AI-based automation can reduce human errors to make forecasts more reliable. At the same time, machine learning (ML) can analyze massive data sets to gain deeper insights and spot patterns indicating emerging risks. Using AI and ML, finance professionals can plan for real-time scenarios, improve operational efficiencies, and improve risk management for greater resilience.
Additionally, AI is now easily accessible. User-friendly AI Chatbots like ChatGPTGemini and Microsoft Copilot have low barriers to entry and are effective for routine tasks such as data retrieval and analysis. There are few real barriers to AI adoption and many potential benefits.
However, there are still many pitfalls that can compromise your ability to realize the promise of AI. If AI isn’t implemented correctly, you risk ending up with confused staff, unreliable information, flawed forecasts, and perhaps even serious security incidents and compliance issues. It’s essential to follow best practices when implementing AI into your FP&A processes, without cutting corners. Here are some tips for successfully adopting AI into your FP&A workflows.
Define an AI strategy
AI is still a shiny new object, but it’s a mistake to blindly rush in and adopt every AI tool you see. Take a step back to establish a cohesive AI strategy before implementing new solutions and processes.
It’s best to start by identifying which aspects of your workflow would benefit most from AI automation or ML analytics. Consult with your finance stakeholders on which workflows should be prioritized for automation and where they find it difficult to gain insight and spot opportunities.
You can then define the tools that will generate the greatest value for your teams. Define the specific benefits you hope to gain from introducing AI, as well as the KPIs and metrics you will track to measure success.
Make sure the databases are strong
AI is not a magic wand. You can’t wave it at fragile data and expect it to generate valuable insights or solve your data analysis problems. The adage “garbage in, garbage out” applies just as much to AI data analysis as it does to manual analysis, and Gartner notes that poor data quality is often cited as one of the main reasons for the slow adoption of AI by finance teams.
It is important to validate your data collection and preprocessing pipelines before introducing AI. Review your data governance policies and ensure there are no silos that could prevent AI tools from accessing the data they need.
Keep humans in the know
Despite all the many benefits of AI, it cannot completely take over FP&A processes. Humans are still needed for several tasks that are not suitable for AI or ML tools. For example, while AI can help with data storytelling, finance professionals need to communicate the insights produced by AI and transform them into a coherent story.
Strategic decision-making is another area that must remain human-led, and finance staff are needed to manage relationships with stakeholders in other departments. Compliance and ethics are areas that are growing in importance as AI becomes the norm and should remain under human management.
Additionally, AI and ML are known to be prone to hallucinations. Human verification of AI and ML results is essential to ensure that crucial financial decisions, predictions and forecasts are not based on faulty assumptions.
Double the security
One of the biggest problems with AI is the question of security. Many finance teams are hesitant to adopt AI solutions, fearing they will harm data privacy or weaken data security. Data security is importantbecause processing large amounts of sensitive information requires robust protection measures. These concerns are also valid: last year, Samsung forbidden employees from using third-party GenAI tools after ChatGPT leaked sensitive data.
International regulations are also catching up with AI and establishing requirements for data privacy and security. It is important to develop clear policies regarding data use, configure and regularly review access permissions, and establish logging and monitoring to track unauthorized uses or access to data.
Consult international best practices on AI data privacy, as they are likely to strongly inform evolving compliance regulations and put their recommendations into practice.
Fostering an AI culture
The world’s best AI tools won’t be of much use if your finance teams avoid using them. Many employees fear that AI will take over their work and/or are distrustful of the technology, leading them to ignore the information provided by AI. Using AI tools effectively also requires digital knowledge and technical skills that your employees may lack.
To overcome this obstacle, invest in creating an AI culture. Reassure your employees that AI is not a threat and present your new solution as a co-pilot that will improve their productivity. Train your employees in the skills you need, although you may need to hire new AI talent. It’s best to start with tools that are user-friendly and intuitive for smooth learning.
You will also need to train finance teams to trust AI. Strive for transparency in AI processes to minimize the “black box” effect. Encourage them to check the AI results first, to help them become familiar with using the results.
AI can transform FP&A – but only with the right approach
It takes careful consideration and rigorous implementation to realize the promise that AI holds to revolutionize FP&A strategies. If you cut corners or ignore the basics, you will get suboptimal results and possibly complete failure. It’s worth investing in laying the groundwork to see your AI solutions succeed.
Learn more
I drive AI strategies for the UK Government, the FCA and PwC. Here’s what they’re doing right – The four approaches common to the most promising AI projects, to help other companies exploit their technology safely and efficiently
Why cloud computing is losing favor – More organizations are moving from hyperscale public cloud computing to multi-cloud and other strategies, says Nick Martindale
The future of quantum computing – what you need to know – Nick Martindale explores the future of quantum computing: how it works, the benefits and the risks you need to be aware of
