Selling CFOs on AI

Remember when Knight Capital's high-speed stock-trading algorithms went haywire back in 2012? The machines, operating outside human control, reacted to market noise and executed millions of errant trades that cost the company a staggering $460 million overnight. A generation of CFOs learned a lesson from that debacle: You can't fully trust technology with money decisions. 

For many CFOs, those lessons have accumulated and been reinforced over the ensuing decade. Today, those with a cautious mindset are reading about the benefits of artificial intelligence but worrying more about the risks. After all, if AI can't tell the difference between a soccer ball and a bald referee's head, or encourages a tech columnist to leave his wife, or hallucinates court cases in legal briefs, how could you trust it with your core financials?

The fact is, AI is ready for CFOs, and has been for some time. AI is already powering a massive range of financial applications on scales at which any fatal errors would already have been uncovered. For example: 

• Transaction processing: Software like Stampli uses AI to automate high-volume accounts payable workflow.
• Reporting: Applications like ThoughtSpot let finance easily access insights from data using natural language AI, instead of using complex SQL queries. 
• Planning: Solutions like Planful apply AI to streamline FP&A, budgeting and forecasting. They surface predictions and recommendations.
• Auditing: AI tools like MindBridge spot anomalies and patterns in 100% of transactions to flag potential fraud. This allows auditors to focus on the outliers.

If you're an advisor to CFOs and finance teams — whether you're a CPA firm, an ERP implementation consultant or even a CFO-for-hire — you know very well that you need to convince your clients to adopt AI. What you probably don't know is how. 

Good news: There's now data that shows you the way. New research on attitudes toward AI explains exactly how to convince skeptical CFOs to change their minds.

The key business benefits of AI

Stampli and Probolsky Research surveyed senior-level finance managers and accounts payable employees in organizations ranging in size from 25 to 15,000 employees. The survey found 89% of senior financial leaders expressed interest in adopting AI. When asked what benefits that expected to receive, it's clear that they have bought in to the hype:

• Improved accuracy (cited by 51%): Humans make errors doing repetitive tasks like transaction coding or data inputs. AI reduces mistakes at scale.
• Faster processing (46%): AI tools work 24/7 executing rules-based tasks faster than humans ever could. This accelerates critical finance processes.
• Enhanced fraud detection (40%): By continuously analyzing 100% of transactions, AI spots anomalies and patterns indicating potential fraud for audit.
• Lower costs (38%): Automating repetitive, manual processes with AI cuts costs substantially. This saves money and frees up finance staff for value-add analysis.
• Increased efficiency (38%): AI tools integrate cleanly with existing financial systems like NetSuite and SAP, increasing workflow efficiency.

When you're speaking to your clients and customers about any AI-based software, it's important to frame its benefits in the context of strategic business impact. 

The biggest barriers

While interest in AI is high, the research also revealed barriers to adoption. When asked about their concerns regarding AI, respondents identified a wide variety, including:

• Data privacy and security fears (cited by 46%). AI depends on data, so CFOs want to ensure it's safe. Stress how your AI vendors exceed regulations, encrypt data and manage access.
• Discomfort with lack of human oversight (41%). CFOs want people involved in decisions. Explain how AI actually augments people with recommendations and control points remain.
• Worry over false or inaccurate information (30%). Hallucination is a concern for all users of AI, but especially in finance, where even a single hallucinated decimal point can spell disaster.
• Replacing jobs (25%). Some worry AI will render finance roles obsolete. This is the case even though AI primarily automates repetitive tasks, allowing finance pros to focus on high-value analysis to guide strategy.

The key insight: human control

The research uncovered a crucial insight for consultants to understand: Across all levels of finance teams, there is a strong preference for humans to remain involved in oversight and control when AI is deployed. 

Specifically:

• 63% indicated they would be less hesitant to use AI if humans had to approve decisions first, rather than AI operating autonomously.
• This sentiment was highest among finance staff closest to daily operations — 64% wanted human sign-off on AI.
• But even 61% of CFOs and 56% of senior finance managers preferred humans stay in the loop to review and approve AI recommendations.

This reveals a powerful lever consultants can pull when positioning AI solutions to cautious finance leaders: Emphasize that the AI software allows humans to stay centered in finance processes with clear oversight of AI. Stress that while the tools provide suggested actions or predictions to finance staff, people still review outputs and have final sign-off before anything is executed.

Be transparent that AI needs monitoring and course-correcting, especially early on. Set the expectation that accuracy and performance improves over time with human guidance. You can overcome reluctance by being clear the tools augment human intelligence, rather than attempt to replicate or replace it. AI and humans each have strengths and weaknesses; your solutions combine both for optimal results.

Additionally, adopt these three best practices:

1. Start small. Pilot a contained AI automation use case like invoice processing to demonstrate benefits. Let success build the case for expansion.
2. Involve finance early. Get user buy-in by including AP teams when assessing AI tools to ensure it maps their workflow needs.
3. Set expectations. Be clear AI won't be perfect immediately. Accuracy improves over time with user feedback and more data. Manage adoption expectations.

Selling game-changing innovations like AI requires education, empathy and patience. Follow these guidelines and you can gradually overcome resistance while delivering major value gains. CFOs will appreciate you guiding them to smarter applications of AI.