Artificial intelligence is changing both sides of payments fraud. Fraudsters can use it to make impersonation attempts more convincing, while finance teams can use it to score transactions, verify documents and identify anomalies.
PYMNTS Intelligence’s “Prevention First: Building a Smarter Defense Against Payments Fraud,” produced in collaboration with Bottomline, surveyed 150 U.S. treasury and finance executives at companies with at least $100 million in annual revenue. Respondents were responsible for or knowledgeable about how their organizations protect outbound payments to vendors and suppliers.
Looking across the findings reveals several divides that help show where corporate payment security is going next.
Here are five things to know.
- Companies Are Spending on Prevention Before AI
Artificial intelligence-driven transaction monitoring ranks high on companies’ investment agendas, but it doesn’t rank first.
Eighty-one percent of firms say they are likely to adopt or improve automated vendor-onboarding verification during the next 12 months, while 75% cite real-time bank account verification before payment. AI-driven transaction monitoring comes third at 73%.
The ordering matters because it shows where companies see the first line of defense. Rather than relying primarily on technology to identify a suspicious transaction after a payment instruction enters the process, firms are putting more emphasis on establishing whether the vendor and destination account are legitimate before money moves.
- AI Adoption Barely Has a Middle Ground
Corporate adoption of AI fraud tools doesn’t follow a gradual progression from one application to several.
Fifty-seven percent of companies use no AI fraud-detection tools, while 42% use three or more. Just 1% report using only one or two.
The distribution suggests that AI fraud detection is being implemented more as a stack than as a standalone capability. Once companies cross the adoption threshold, they tend to combine applications such as real-time risk scoring, document verification and scanning communications for AI-generated markers.
That divide could narrow quickly. Among companies that haven’t yet adopted AI fraud detection, 24% are implementing it now, 9% expect to within six months and another 25% within 12 months.
- False Positives Are a Bigger AI Problem Than Doubts About Effectiveness
Companies on both sides of the AI adoption divide share a concern: stopping legitimate payments.
Among firms that don’t use AI fraud detection, 89% cite fear of flagging legitimate payments as an adoption barrier, while 74% are hesitant to automate without human sign-off.
Experience with the technology doesn’t eliminate those concerns. Among AI adopters, 92% say concerns about incorrectly flagging legitimate payments made adoption difficult, and 89% cite hesitation about automating without human approval. Cost is cited by 49% and legacy-system difficulties by 32%.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
The data puts governance and accuracy alongside detection performance as core implementation issues. Companies aren’t only deciding whether AI can find fraud; they also have to determine when a machine-generated warning should interrupt a legitimate payment and when a person should make the final call.
- Goods Companies Have the Widest Gap to Close
Industry comparisons expose a particularly sharp divide among companies selling physical goods.
Just 34% of goods-sector firms use three or more AI fraud-detection tools, compared with 44% of services companies and 57% of technology firms. Goods companies also have the lowest share stopping at least 90% of fraud attempts before a loss, at 41%.
At the same time, 54% of goods companies without AI are satisfied that manual processes handle fraud well enough, compared with 33% of technology firms.
Yet goods companies that have adopted AI use it heavily: 90% of those adopters use real-time risk scoring and 90% scan emails and texts for AI-generated markers.
The divide, then, is less about how aggressively adopters use the technology than about how many companies have crossed into adoption at all.
- Some Fraud Exposure Still Comes Down to How Companies Pay
The AI discussion can obscure an older source of payment risk: checks.
Thirty-five percent of surveyed firms report encountering check fraud. Among all firms, 48% say checks account for at least 10% of outbound fraud losses, including 15% that attribute more than one-quarter of their losses to checks.
AI can improve detection around a payment, but the payment method itself remains part of the risk equation.
That gives payment modernization a security dimension alongside its familiar speed, cost and automation benefits. For finance teams evaluating fraud investments, changing how money leaves the business can sit alongside improving the technology watching it.
At PYMNTS Intelligence, we work with businesses to uncover insights that fuel intelligent, data-driven discussions on changing customer expectations, a more connected economy and the strategic shifts necessary to achieve outcomes. With rigorous research methodologies and unwavering commitment to objective quality, we offer trusted data to grow your business. As our partner, you’ll have access to our diverse team of PhDs, researchers, data analysts, number crunchers, subject matter veterans and editorial experts.