Questions to keep AI strategy on track

Vin Kumar and John O’Mahony at The Hackett Group pose six questions that CFOs should ask about their AI strategy

 

It’s often impossible to know in advance which aspects of a new consumer technology will take off and which will be tomorrow’s eight-track tape. 

 

The same is true for business-to-business technologies – up to a point. With an operational technology, you may not be able to predict where something is heading or foresee the full extent of its ramifications.

 

But you can generally see it coming – think for instance about the Internet of Things or big data. Artificial intelligence (AI) and generative AI (Gen AI) are other big things that belong in this category. Even taking the hype into account, AI is maturing rapidly and becoming an integral part of a wide range of both consumer and business systems.

 

Most of the time, such a big-picture view would be enough for the chief financial officer (CFO) and their finance team to go on. However, AI is special. It’s changing – and will continue to change – how value is generated within your company, and how your company’s activities are recorded.

 

The range of potential impact is so wide that the finance team will need to maintain a clear sense of where the technology is today, where it is headed, and where and when should we adopt it.

 

Here are six questions to ask now: 

 

1.      How is our software changing? 

Over this next year, Gen AI is likely to be integrated into most of the software solutions that you already use, giving you access to some powerful processing power without the need to do anything.

 

However, you should reserve some time for your team to learn how to take advantage of its capabilities, decide what you want to leverage and what you need to build, and help optimize working practices.  

 

2.   What are our vendors doing with AI? 

Reach out to find out what AI capabilities your vendors – both solution and service providers – are planning to incorporate to assess which aspects of their expertise you should tap into and where you may need to build your own AI solutions.  

 

Begin by talking to your professional service partners about any plans they may have to roll out AI-enhanced services. In the second half of 2024, we expect to see specialized large language models enter the market – systems that will offer analytic support for areas such as internal auditing, external reporting and tax reporting. 

 

You will also want to talk to the more general business process outsourcers you work with and ask them if they have specific plans to roll out AI-enhanced services. The goal is to make sure that you aren’t duplicating their efforts. Knowing what they are thinking now should give you some important insights into where your department should begin to focus on enhancing value and complementing where your external partners are investing in AI.

 

3. Where should we focus AI power first?

AI is maturing rapidly, and its capabilities are constantly expanding. When requests for pilot funding comes in, you need to keep pushing back until someone can articulate a clear and tangible business case based on things that the tools chosen can actually do.

 

You will only want to give your blessing to pilots that have the ingredients of a good proof of concept: value enhancing, the right technology, right data and right talent on hand to execute the project. You should identify pilot use cases to use AI as both a digital assistant and also to integrate it so that it enables your process automation initiatives.

 

4. Are we considering all automated sourcing opportunities?

It’s early days for a lot of business applications of Gen AI, but this is significantly less true in the area of procurement, where automated sourcing systems are already saving companies millions. Such automated sourcing systems as Globality, Fairmarkit and Pactum are taking over important aspects of procurement, including key tasks such as market intelligence, negotiation, contract drafting and reviewing.

 

While these opportunities are likely to be dwarfed later by advances in other industries – one recent study by The Hackett Group estimates that selling, general and administrative costs may be slashed by as much as 40% when Gen AI matures in five to seven years – procurement is one area where the technology is pulling its weight today.

 

5. Are we protected against AI legal risks?

In the past, computer programs could be counted on most of the time to produce answers that were more or less correct. Now, the probabilistic nature of Gen AI makes it a front for creative answers that might appear correct, but are actually “hallucinations,” which basically means they are fiction – not fact.

 

At least for now, it seems that models that are built out of large lakes of data are prone to some of the same errors as the human brain. But these models make those mistakes more quickly and at a much larger scale. Are you protected if one of these models goes wrong and harms a customer? Does the technology create any regulatory risks? What happens if your financial data leaks out? Who profits from intellectual property you develop in cooperation with a third-party AI developer?

 

Your legal team should consider some of these issues now and begin incorporating these risks into their contracts. If they wait until a crisis arrives, it may be too late.

 

6. Who should be on the CFO’s AI team?

Certain kinds of AI, such as robotic process automation and machine learning, have been around for a decade or more. Others (particularly Gen AI) are just taking shape right now. When it comes to staffing for finance-related AI initiatives in 2024, this means that you need to focus more on staff who have enthusiasm and curiosity than for expertise. 

 

Once you have found candidates you believe in, encourage them to absorb as much as they can about AI, particularly Gen AI, which operates on quite different principles than earlier varieties. Give these team members the time and resources they need to get up to speed with the technology today.

 

One good place to begin: have them sign them up for some of the free courses on Gen AI basics offered by Amazon, Microsoft, Google and other major tech companies.

 

With a technology that is so young and evolving so quickly as AI, it’s going to be tough to get every answer right. Fortunately, asking good questions should help keep you on track.

 

---

 

Vin Kumar is AI and Global Digital Operations Practice Leader at The Hackett Group and John O’ Mahony is their Head of Finance and Business Transformation

 

Main image courtesy of iStockPhoto.com