By Rachel Burger February 18, 2025
There’s no doubt AI is a game-changer for FP&A teams.
And increasingly it’s becoming a necessity. OneStream’s recently released survey Finance 2035: Return to Investment revealed that 70 percent of CEOs and 68 percent of CFOs believe that organizations that fail to invest now in AI tech, infrastructure, and skills will not survive the next five years.
Yet while AI offers a path to unprecedented efficiency and innovation, it also raises new questions and potential ethical concerns if it is not implemented and used in thoughtful and disciplined ways.
Developing an ethical framework and protocols for the use of AI not only builds trust within your organization and with key stakeholders but also creates safeguards from unintended consequences or unnecessary risks.
When developing an ethical framework for AI, the first step is to embrace an overarching mindset that views AI as a powerful tool, but not a replacement for human judgment and oversight.
There are countless, nuanced decisions that rise well above AI’s pay grade, and require human insight and collaboration to provide guidance that stays true to ethical practices.
That mindset includes finding a balance between automation and oversight in ways that ensure that you can reap the full benefits of AI while minimizing downside risk. It’s key to make sure your team and any relevant stakeholders receive training in AI ethics and that protocols are put into place for human review of AI recommendations that may venture into ethical gray areas.
With that in mind, here are key areas where ethical considerations could come into play.
Avoiding potential bias
AI helps create real-time efficiencies and powerful forecasting capabilities, often by relying on historical data to make recommendations and predictions.
In some instances, that can introduce the possibility that historical data includes biases that could emanate from long-standing inequalities, demographic imbalances, or flawed or incomplete data collection.
To limit this risk, make sure training datasets are diverse and representative and regularly audit AI models and outputs for potential bias. Adding diversity to your team or developing an AI advisory board that includes diverse individuals from within your organization can help spot potential blind spots before they become a major issue.
Increasing transparency
It’s impossible for AI systems to account for the many potential complexities and nuances that go into many decisions. For FP&A, where maintaining trust is essential, relying too heavily on AI can limit transparency which can lead to mistrust and skepticism. Stakeholders need to understand why specific recommendations are made, especially for high-stakes decisions.
To enhance transparency, make sure your finance team has input into the how and why AI is being integrated into your workflow, and are able to clearly explain why decisions regarding AI deployment and functions were made.
Ensuring data privacy and security
AI can definitely be your friend when it comes to data security by automatically detecting anomalies or irregularities in financial transactions in ways that quickly identify potential fraud, errors, or inefficiencies.
Yet the fact that AI-driven FP&A tools require access to sensitive financial data and information also can introduce increased risk of unauthorized access, misuse, or data breaches that can lead to costly repercussions for your organization.
To uphold data privacy and security, make sure AI systems have robust encryption and cybersecurity measures and comply with relevant data protection regulations, such as GDPR or CCPA. By building security protocols into your AI integration and operation plans you can help ensure that you keep one step ahead of the hackers.
Creating Accountability
When AI systems contribute to errors, determining accountability can prove challenging. For example, if an AI tool introduces bias or makes a recommendation that leads to action that doesn’t align with your corporate values, the question becomes who is responsible?
Ideally, accountability protocols and guidelines are established early in your AI journey to provide roles and responsibilities for all stakeholders are clearly defined, and there are mechanisms in place for dispute resolutions. Again, ensuring ongoing and vigilant human oversight can help avoid problems before they occur.
Executing ethical investment practices
Using AI to help make investment decisions might prioritize maximizing profit without considering ethical implications. For example, an AI could recommend investments in industries with that some stakeholders or customers could view as having negative social or environmental impacts, such as fossil fuels or arms manufacturing.
To help ensure ethical investment practices, incorporate Environmental, Social, and Governance (ESG) criteria into AI algorithms and make sure there is clarity and consensus around your organization's values and standards.
Considering AI for your finance team? Take our AI readiness assessment.
