Rethinking Finance With an AI-First Approach

Layering AI onto existing operations might boost short-term efficiencies. However, a piecemeal application of the technology delivers incremental gains.

What if, instead of lopping off inefficiencies at the edges, CFOs could completely redesign finance operations from the ground up, with AI at the core and boost gains on new levels that go beyond productivity? Such a tantalizing possibility is well within reach with an AI-first approach to finance. Consider the example of the financial close cycle time. Instead of waiting for the end of a financial period, switching to continuous close gives finance teams the information they need at their fingertips all the time, which can result in faster decision-making. This reimagination of everyday business operations is precisely what AI-first finance delivers. In a rapidly evolving digital landscape, the very survival of business demands a strategic reboot of the AI mindset. And AI-first finance isn’t about blindly using the technology in each and every process.

Instead, it’s about evaluating finance operations with an AI lens and using the technology when and where it can deliver radical improvements.

But making this transition isn’t without its challenges.

Adoption Barriers and Talent Shortages

Related:How Embracing Data Fabric Can Enhance the Value of AI Initiatives

CFOs who want to adopt an AI-first approach might contend with a few stumbling blocks:

Indeed, according to a study done in partnership with HFS Research of 550 senior executives on their generative AI adoption, respondents from finance functions say that “difficulty in finding, training and retaining AI talent” (34%) and “lack of data quality or strategy” (34%) are two of the top three barriers to the adoption of gen AI. Finance leaders are tackling these challenges head-on.

To address the talent gap, for example, they’re investing in upskilling in-house talent, bulking up research and development departments and forging partnerships with specialty providers.

In addition, 40% of finance executives in the gen AI survey say they’re collaborating with technology providers specializing in gen AI solutions as a way of building and enhancing their own technology capabilities.

Once these obstacles are addressed, CFOs can start putting their AI-first approach into action by focusing on four key aspects.

Related:The Future of AI Talent Acquisition in Europe

The Four Building Blocks of AI-first Finance

1. Data empowerment

AI thrives on data that is accessible and discoverable. However, without a single source of truth, data can become siloed, leading to multiple versions that are out of reach for even internal departments in the enterprise.

This is why prioritizing data management is a must.

Both unstructured and structured data needs to parse and route to the right systems in a continuous loop. Because AI can quickly tap into accurate data from across the business, freight accruals, for example, can now be accounted for on a shipment basis as soon as goods leave the warehouse, rather than monthly, with adjustments for leakages on goods received versus invoices received.

Additionally, in the past, the finance team would have to run SQL queries or specialized database commands to get the data they needed. Now, they can simply type a request in everyday language and retrieve information much more easily, increasing data accessibility.

Underpinning these developments, it’s critical to establish solid business practices for master data governance, ethics and compliance. Otherwise, companies may be at risk for legal and regulatory issues, data breaches and reputational damage.

2. Technology that scales

A robust technical architecture makes a key difference in AI-first finance. An optimal tech stack includes four essential layers:

When looking at the four layers, almost all enterprises will have an SOR, but the SOO is most likely to have been overlooked.

An AI-first approach blurs the lines between these four layers by helping each layer to work smarter and more seamlessly together.

For instance, Copilot in Microsoft Excel can automatically handle transaction matching and financial reconciliations. It can also send follow-up emails or messages directly from the spreadsheet and manage workflows using Power Apps, a separate application—a prime example of SOE, SOI and SOO all coming together.

3. Algorithm-driven operations

In AI-first finance, every finance operation boils down to an algorithm, a set of rules or instructions, that integrates into the larger tech ecosystem. Every business challenge can be turned into a predictive task that algorithms can address by forecasting solutions.

For example, warranty accruals involve estimating the number of claims a company expects to receive. By predicting the likely number of warranty claims, the company can set aside appropriate funds. These accruals are then recorded in the financial statements at the end of the month.

CFOs must embed AI-powered algorithms to improve finance operations and outcomes.

4. Human oversight for responsible AI

Results from AI are not binary. When AI offers an answer that gives a confidence value of 94%, it means it needs oversight or improvement by people with the right knowledge and context. Having humans in the loop is as important as ever, regardless of confidence score—especially in finance, where faulty or biased AI decisions can lead to severe outcomes. 

As much as enterprises work to develop bias-free algorithms, humans help make sure AI-based decisions reflect the values of the company and are equitable, just and accurate. We must look at the impact of AI on the existing control environment and consider what shifts are needed, as well as the ways those controls are embedded in different processes.   

Unlocking Exceptional Outcomes

AI-first finance is already reimagining established practices. Zero-touch processing with zero time to close, zero time to insights and zero exceptions are entirely within sight. And reconciling monthly statements can be a thing of the past for enterprises that have rethought their close cycle times with AI. 

By focusing on continuous learning, experimentation and adaption companies can harness AI’s full potential by embedding it into finance operations, processes and decisions.

And those that do are realizing unparalleled results.