The scale of financial fraud is putting renewed attention on how quickly financial institutions and companies can identify problems before they become losses. The FBI’s newly released 2025 Internet Crime Report, published earlier this week, recorded more than one million complaints and nearly $21 billion in losses, up significantly from the previous year. Cyber-enabled fraud accounted for the overwhelming majority of those losses.

The figures reflect a wider challenge for the accounting and finance profession. As financial activity becomes more digital and interconnected, detecting problems only after a transaction or reporting cycle has been completed can be costly. Consequently, more attention is being placed on controls that identify unusual activity earlier, before unreliable information or suspicious transactions move further through financial systems.

One example of this earlier-control approach is the work of accounting and finance professional Elisha Adeboye, whose experience across corporate reporting, audit and financial technology has increasingly focused on identifying financial problems before their consequences spread.

The need to identify financial problems before they become deeper-seated in reporting systems is particularly important within large multinational companies, where financial information passes through multiple entities and reporting layers before reaching management, investors, and regulators. At that scale, effective financial controls depend heavily on identifying and resolving unusual movements before they flow into wider corporate reporting.

At Cummins, where Adeboye worked in corporate accounting and external reporting, the challenge was the sheer complexity of a global business: large currency exposures, joint ventures, acquisitions, and financial information moving across multiple entities before reaching consolidated reports.

That environment gave him a close view of how quickly an unexplained number can become a much bigger reporting problem. A Cummins finance leader who worked closely with Adeboye described his involvement in reviewing unusual joint-venture movements and identifying a discrepancy that could have inflated reported figures by about $5 million.

He had seen the same problem from a different angle at PwC Nigeria. A former PwC partner who directly supervised Adeboye described one audit in which his analysis identified a revenue-recognition issue worth more than $2.5 million, alongside other reporting questions that required looking beyond whether the numbers merely added up.

What connected those experiences was timing. Identifying an error after a report is issued is useful; identifying it before it travels through a company’s reporting chain is far more valuable. That thinking followed Adeboye into financial technology. A colleague who worked with him at Intuit described his involvement in developing a financial-management platform for small businesses, bringing a similar emphasis on visibility and early detection closer to everyday users.

His more recent work pushes that idea further. Adeboye developed a fraud-prevention system that uses behavioral biometrics to detect unusual activity even after someone has successfully logged in to an account. Instead of relying solely on a password or one-time code, the system monitors changes in how a person interacts with a device. A sudden departure from established behavior can trigger another check before a suspicious session turns into a financial loss.

That is the broader shift Adeboye’s work reflects: financial controls are moving closer to the point where something first starts to go wrong, rather than waiting for the damage to appear later in a ledger, an audit or a fraud report.