Many finance teams are optimistic about the potential impact of adopting AI in their workflows. However, concerns remain, particularly regarding data integrity, privacy and skill gaps.
These were among the key insights uncovered in our 2025 State of Strategic Finance Report in collaboration with BPM Partners.
At Excelerate Finance 2025, John Colbert, Vice-President of Advisory Services at BPM Partners, shared some useful tips during his session, “Strategies in Driving AI Adoption in a Modern Finance Organization,” to help finance teams navigate these concerns.
In this article, we’ll break down what he shared, everything from the evolution of AI adoption in finance to real risks you shouldn’t ignore, and practical strategies to address them and roll out AI in a way that feels manageable and low-risk.
With over 30 years of experience in the software industry, John Colbert has witnessed firsthand the evolution of financial technology, from early computerization to today’s AI-powered tools.
In his words, this shift has been decades in the making. It didn’t start with ChatGPT or Microsoft Copilot. Rather, it began in the 1980s with the rise of computerization. In the 2000s, business intelligence and reporting tools gained traction. By the 2010s, predictive analytics had begun to play a more central role in finance teams.
Today, AI and machine learning are no longer fringe concepts; they are becoming integral to day-to-day finance conversations, and the mindset surrounding them is changing rapidly.
“This shift in attitude is noteworthy. If we rewind to a year ago, most finance professionals, understandably, were much more conservative about AI. But that is changing fast. Nearly 60 percent now say they are using it in some form. Now is the time to move from dipping your toes in the water to getting your feet, and even your knees, wet. It is about deepening adoption and growing your understanding of how these tools can serve your team.”
John Colbert, VP of Advisory Services, BPM Partners
The most common use cases cited by finance leaders, according to our State of Strategic Finance 2025 report, include data analytics, predictive modeling, anomaly detection, and generative capabilities.
With finance teams playing an increasingly strategic role in business planning and collaborating more with IT, Sales, HR, and Operations, John notes that AI has the potential to strengthen that role even further.
The survey responses back up this fact. Eighty-two percent said they feel optimistic about AI’s capabilities and potential impact for their department.
While acceptance of AI is picking up, there are still real risks that come with adopting AI in finance.
Based on his experience gathering insights from businesses about their use of AI, John outlined five common risks that finance teams should plan for, along with ways to manage them before they slow down adoption or create new challenges.
This is one of the most common challenges finance teams face, and it is not new. Even without AI, working with messy, siloed, or hard-to-access data slows everything down. But when AI enters the picture, the risk gets even higher.
AI tools for finance need clean, reliable data to deliver useful insights. But if the data is scattered or incomplete, the output quickly loses value. It becomes harder to trust the results or spot important trends.
This is not just a theoretical concern; it is a reality that teams face every day.
In our State of Strategic Finance 2025 report, nearly 40 percent of respondents said that accessing data from multiple sources remains one of their biggest challenges, a concern that has persisted since our 2022 report.
John encourages teams to tackle the challenge of data quality before introducing AI into the picture, as it pays long-term dividends.
Start by investing in better-integrated tools and improving your extract, transform, and load (ETL) pipelines. This ensures that data flows smoothly between platforms, rather than staying locked in disconnected spreadsheets or legacy systems.
John also highlights the need for strong data governance.
“Make sure you have strong data governance within your finance or FP&A teams. You need clear protocols and guardrails at the input and consolidation stages to keep your data clean. AI models perform better with larger volumes of data, but you still need to structure that data in a way that makes it usable, even without AI.”
John Colbert, VP of Advisory Services, BPM Partners
Simply put, more data isn't always better. If your data is messy or poorly structured, even the smartest AI tools will struggle to deliver useful insights.
Despite the buzz surrounding AI, many finance professionals remain hesitant to fully embrace it. Some of that hesitation stems from a lack of trust or fear of losing control. And it makes a lot of sense.
Many finance teams worry about not being able to trace how AI outputs are generated. Without a clear audit trail, it becomes harder to validate results, ensure compliance or explain outcomes to stakeholders.
For teams that are skeptical or unsure about AI, John recommends starting small and training your team to view AI as a support tool, not a replacement for the hard work that finance teams already do and need to continue doing.
In other words, begin experimenting with manageable use cases, especially features that are already being introduced into tools your teams use every day, like Excel. This lowers the barrier to entry and helps people build trust through firsthand experience.
“Whether you actively adopt AI or not, you’re likely already seeing it show up in your Excel models and in the tools you use every day. See it as an opportunity to learn more and build trust in these systems.”
John Colbert, VP of Advisory Services, BPM Partners
Once small experiments begin to show value, it’s also helpful to share those early wins internally. When colleagues hear that someone used AI to speed up forecasting or automate a recurring report, it becomes easier to see its practical benefits.
John shared a customer example where internal demand around AI grew organically as people saw the benefits and began asking for access:
Englobe, a professional services firm with around 3,000 employees, was expanding through acquisitions, which brought new challenges. During their last budget cycle, they worked with about 500 linked spreadsheets.
They had already been testing Microsoft Copilot and other AI tools before trying Vena Copilot in April 2025. Their biggest concerns at the time were the speed of adoption, security, and the risk of hallucinated outputs, all legitimate worries that have historically slowed AI adoption for finance teams. To address these, they took a cautious, phased approach.
First, they focused on gaining regional buy-in and encouraged adoption from the ground up, rather than forcing it top-down. They also appointed a dedicated AI project lead whose job was to manage feedback, share updates, handle internal communication, and check in regularly.
This way, existing team members felt supported, as they didn’t have to carry the burden of building the AI expertise all by themselves.
Francis Paquette, Director of FP&A at Englobe, offers this advice to teams considering implementing AI:
“AI is going to do some amazing things for us, but start low. Set expectations low, and make sure you exceed them.”
Englobe chose to build on what was already working. After investing in their Vena setup, they used those existing models as the foundation for their AI rollout, leveraging previous wins to ease adoption and grow internal confidence.
Finance professionals aren’t always experts in deep data analysis, data science, or AI. In fact, these were among the largest skill gaps reported by survey respondents in our report, with approximately one-third citing data analysis and 20 percent citing AI/machine learning as the weakest areas of their team.
These are real barriers, but they also present a clear opportunity for finance teams looking to adopt AI in their workflows.
As John suggests, finance leaders can:
Launch cross-training programs to gradually upskill their teams
Partner with platforms like Vena that offer guidance on adopting AI through hands-on training, educational resources, and community support
Explore low-code or no-code tools that reduce reliance on IT and make it easier for non-technical users to get started
Prioritize tools built specifically for business users (using natural language prompts as the entry point) to lower the learning curve and speed up adoption
Build internal partnerships with IT or data teams to create the right support structure for adoption
Michael Ragsdale, VP of Finance, Strategy and Analytics at the Kansas City Chiefs, shared during his session at Excelerate Finance 2025 that his lean finance team is supported by four data scientists. This collaboration helps the finance team manage advanced modeling and automation without having to learn every technical skill themselves.
One of the key risks finance teams face when adopting AI is poor integration between new AI tools and their existing systems. This kind of disconnect creates data silos, broken workflows, and unreliable AI outputs. It can also slow down adoption, reduce trust in the results, and increase manual work, undermining the efficiency AI is meant to provide.
For example, using standalone tools like ChatGPT outside your core systems can make it harder to connect outputs to your source data. There’s also the risk of sensitive inputs being used to train public models if settings haven’t been properly adjusted. That’s a serious concern for finance teams who work with confidential data by default.
John suggests that finance leaders:
Choose AI tools that work within your current planning stack instead of relying on third-party apps that don’t sync well
Audit your current systems to identify integration gaps before adopting new tools
Set up clear guardrails to manage which tools are introduced and how they access your data (your team might already be using AI without your knowledge)
Prioritize platforms with trusted data models, so your team can have confidence in the AI outputs from the start
When selecting AI tools, it’s essential to consider how well they integrate with your existing technology stack. Tools that connect seamlessly with your ERP or planning software are more likely to support smoother, more consistent workflows.
If your team is already standardized on Microsoft’s ecosystem, you’re in a strong position to take advantage of tools like Microsoft Copilot and Vena (which is built to help you maximize your existing investment in the Microsoft suite of tools).
Vena Copilot’s integration with Microsoft Teams, for example, helps teams to embed AI insights directly into their workflows and share them across the business. This makes it easier to embed AI into day-to-day work, rather than treating it as a separate tool.
Bottom line? The more aligned your systems are, the easier it becomes to scale AI adoption without compromising data accuracy or operational control.
While AI can speed up analysis and surface insights quickly, it can also oversimplify complex issues or misinterpret context-sensitive decisions. That is why human judgment must remain central in financial planning.
John recommends that finance leaders stay actively involved and treat artificial intelligence as a value add, not a fallback. For the best results,
Maintain a human-in-the-loop process for reviewing AI-generated outputs
Set up escalation protocols for low confidence scores or questionable results
Use AI to support analysis or build arguments, but always validate results manually
Train teams to understand where AI insights come from and what methodology was used
These steps help ensure that AI becomes a powerful support layer, not a black box.
“Finance leaders should never be caught off guard when asked, ‘Where did this number come from?’ or “Why is this report saying this?” You cannot respond with, ‘The AI generated it.’”
John Colbert, VP of Advisory Services, BPM Partners
So, while AI can assist the process, leaders must still understand the context behind the numbers to stand behind them. The same applies to junior team members, who may be tempted to rely on AI without fully understanding the business context. John advises managers to challenge them in meetings, asking:
Where did the data come from?
What does it mean?
Why does it look this way?
Drilling into the data helps your team build the critical thinking skills they need and makes AI a tool for growth, not a shortcut. This learning process is easier when teams use tools that make it simple to trace the source data and the logic behind AI-driven outputs.
Vena Copilot in particular, shows exactly how it arrived at a result, allows users to provide feedback on accuracy, and lets teams tailor rules to their business. Its auditability and model training features help newer team members build trust in the numbers while working within proven models the organization already uses.
A view of how Vena Copilot explains the methodology behind its answers.
To adopt AI successfully, John suggests conducting a self-audit first.
Ask yourself honestly: Are we leading, advanced, or still developing when it comes to AI readiness? Are we equipped to adopt right now, or do we need to lay more groundwork? These are essential questions for any finance leader.
He also advises taking time to put the right controls in place and rolling out AI in a way that reflects your team’s maturity, the sophistication of your tech stack, and the models you’re building from your data.
AI adoption doesn’t have to be a giant leap. It can be a series of small, strategic moves, and the real risk is falling behind. So, start with one focused, low-risk use case. From there, learn, refine, and scale based on what works best for your team.
Want to dive deeper into how finance teams are embracing AI? Read the full State of Strategic Finance 2025 Report.
Learn practical strategies for managing the complexities of AI adoption with John Colbert, VP of Advisory Services at BPM Partners.
Watch NowJessica Tee Orika-Owunna is a content strategist and writer with over seven years of experience creating and repurposing relatable, helpful content for global brands including Contentsquare, Softr, Hotjar and Vena. She specializes in turning everyday product, user, and subject matter expert insights into product-led content that answers real buyer questions and supports better business outcomes.
