Companies lose more than 5% of their revenue each year to fraud, according to data from the Association of Certified Fraud Examiners (ACFE). The tactics deployed by fraudsters are often surprisingly mundane. They are subtle, sometimes hidden among routine payments or ordinary payroll processes and easily mistaken for human error.
Financial fraud therefore can go undetected for many months. Just last week, a former KPMG consultant was charged with siphoning nearly $1m (£700,000) through fraudulent tax returns using unauthorised credentials. He had lodged false claims for years before being detected.
Keith Williamson, managing director of the disputes and investigations team at Alvarez & Marsal, a consultancy, has worked as a forensic accountant for more than 25 years. He says the sheer volume of transactions makes fraud difficult to detect. “The number of transactions recorded by companies is phenomenal,” he says. “Compared to the past, when records were handwritten in ledgers, it’s much easier now for illicit activities to blend in. It’s also become much easier to forge documentation and data.”
A domineering CFO whose decisions go unchallenged creates an environment ripe for misconduct
One in three finance professionals say detecting fraud is among their top challenges, according to a survey by Medius, a spend-management firm. And the rise of AI may be exacerbating the problem. Two in five (42%) respondents suspect their colleagues have submitted receipts that have been falsified or altered by AI.
Williamson highlights three main categories of fraud: management, employee and financial misstatement. The motives behind them vary, but economic pressure is often the primary driver. “Companies may manipulate results to side-step breaching bank covenants, prevent loan recalls or avoid higher interest charges,” he says. “Sometimes the goal is to prop up share prices – a powerful incentive when senior management holds significant stock in a publicly traded business.”
In other cases, fraud is purely opportunistic. An employee seeking personal gain might exploit weak controls or take advantage of lapses in oversight to commit theft or embezzlement. “These individuals often rationalise their actions as temporary or justified, but small infractions can quickly escalate into significant losses if unchecked,” Williamson says.
Spotting fraud as it occurs requires finance teams to detect its patterns and understand the motives behind it. Here, three forensic accountants reveal the red flags often buried deep in company ledgers.
Atypical outflows that seem normal
For firms with high transaction volumes, fraudulent activity can be easily hidden in plain sight. The sheer number of transactions provides the perfect camouflage, allowing illicit payments to slip through routine processing with minimal review.
Suspicious signs include repeated claims falling just below approval limits, duplicate payments to the same vendor month after month, increasingly frequent ‘miscellaneous’ payments and reimbursements lacking supporting documentation.
Payroll isn’t adding up
“Payroll fraud is one of the most common forms of financial misconduct,” says Williamson. According to the ACFE, it goes undetected for 18 months on average – during which time significant financial damage can occur.
Payroll schemes range from falsified timesheets to ‘ghost employees’, where made-up workers quietly draw salaries, Williamson explains.
Finance teams should look out for time records that do not match operational logs, payroll increases without corresponding headcount or wage growth, payments to terminated or fictitious employees or department expenses that are misaligned with current staffing levels.
Back-dated and unexplained journal entries
Frequent manual or back-dated adjustments with no explanation are also signs of potential accounting manipulation.
Such entries may be created with no underlying transactions to support them and simply exist to distort the financials, says Martin Chapman, national head of forensic services at Crowe, a UK-based accountancy firm. “Every journal should relate to a genuine accounting event. If not, it can indicate manual alteration and manipulation of the accounts.”
Records of payments that are likely but not certain are especially vulnerable to manipulation. For instance, teams may exaggerate provisions for potential legal damages in an ongoing lawsuit or record revenue from a long-term contract before it has been earned and processed.
Vendor spoofing
Fraudsters can exploit trusted relationships with vendors by setting up fake companies whose names resemble legitimate partners. If an approver skims a payment request and sees a familiar-looking name, they will often authorise the request with little scrutiny.
Finance teams should be alert to even minor discrepancies in suppliers’ names, vendor addresses or tax IDs that match employee records and new payments made to vendors whose relationship with the company had previously been terminated.
‘Friendly’ transactions
Related-party transactions are deals between entities with some connection before the transaction. For instance, two companies with a shareholder in common are related parties. Although they can be legitimate, there are many restrictions on such dealings in the UK and they are frequently used for financial manipulation. Related-party transactions can be used to boost sales figures, for instance.
“Auditors struggle to spot this type of fraud because they depend on the organisation to fully disclose all related parties,” says Chapman. “Even when firms try to be transparent, vague accounting standards make it challenging to identify every related party. On top of that, management doesn’t always understand the risks involved, leaving gaps where fraud can go unnoticed.”
Finance teams must be aware of all commercial relationships across the business and ensure that any transactions with ‘friendly’ third parties are recorded properly.
Mismatches between operations and finances
Another sign of fraud or manipulation is misalignment between the records in a company’s operational systems, including sales reports, inventory counts or production data, and the figures in its financial statements.
Even small inconsistencies, such as a slight irregularity in inventory turnover, should be investigated. That’s according to Mihaela Ceornoava, a financial controller at Omniconvert, a marketing firm. Comparing the real-world data with what is recorded in the financial books can uncover discrepancies that a standard audit might miss.
“Don’t ignore small mismatches between operations and finances; they can be the thread that unravels a bigger problem,” she says. “Keeping detailed, well-structured records of these anomalies turns a gut feeling into solid, defensible evidence if fraud is suspected.”
Siloed accounting duties
Granting too much financial control to just one individual greatly increases the risk of fraud. If one person, or even a small team, is given responsibility over multiple financial duties, such as authorising payments, managing vendor records and reconciling accounts, they could easily perpetrate a fraud with little or no collusion.
“This risk is especially pronounced in fast-growing businesses that have not invested in strong internal controls or segregated duties,” Williamson says. “Without proper checks and balances, errors or manipulation can go unnoticed for months or even years, leading to significant financial losses and reputational damage.”
By implementing clear separation of responsibilities, mandatory dual approvals and regular independent audits, finance chiefs can mitigate this vulnerability and protect the organisation against costly fraud.
Attitude of ‘it’s just a small amount’
Not all fraud starts with grand intent. Many cases begin with small, opportunistic acts, such as minor unauthorised transactions that snowball when no one notices or acts. Weak oversight and lax controls create an environment where one-off moments of misconduct quietly escalate into major operations.
“Brushing off concerns because ‘it’s just a small amount’ or ‘that’s how it’s always been’ indicates poor financial controls,” warns Williamson. “Similarly, a domineering leader or CFO whose decisions go unchallenged creates an environment ripe for misconduct.”
Organisations will be vulnerable to exploitation if its processes to prevent and detect fraud and accounting manipulation are inadequate. Finance teams should conduct scenario-planning and training on common risks to ensure frontline staff are well-equipped to identify and report any red flags.
Very little human oversight
“An absence of human judgement is a critical warning sign,” Williamson says. “Professional expertise is needed to interpret anomalies, question suspicious patterns and prevent AI-generated fraud from going unnoticed.”
Advancements in generative AI models are making fraudulent claims harder to detect and easier to produce. One in three finance professionals would fail to spot an AI-fabricated receipt if it landed on their desk, according to the Medius survey. Another 30% reported a rise in falsified expenses since last year.
While some finance teams have adopted AI-powered systems that flag subtle warning signs, such as formatting inconsistencies, duplicate vendor documents or unusual transaction metadata, technology alone isn’t enough to solve the problem, Williamson stresses. “Without strong human oversight, critical context and judgement are lost.”
Finance teams should apply professional scepticism to every review, watching for unusual visual cues such as mismatched dates, inconsistent fonts or misplaced logos. Firms could also introduce verification procedures to authenticate documentation before payments are approved.
Employees not taking breaks
Employees who consistently avoid taking annual leave could be up to no good. “While this might look like dedication on the surface, in finance, it should actually raise concerns,” explains Ceornoava. “People who commit fraud often need to stay close to the books to keep things hidden. These schemes require constant oversight to conceal irregularities, meaning the perpetrator can’t risk someone else stepping into their role.”
Enforcing mandatory breaks and rotating duties is not only important for wellbeing, it can also help to bring any financial discrepancies or unauthorised transactions to the surface.
“Often, organisations adopt a reactive rather than proactive response to tackling fraud, which means red flags can be missed that would have otherwise been picked up if the right controls and processes were in place,” says Chapman.
Best practices for early fraud detection
Conduct quarterly mini-audits on high-risk processes, such as procurement or related-party transactions.
Create a cheat sheet of financial (e.g. unusual journal entries near reporting deadlines) and non-financial (e.g. changes in supplier ownership without disclosure) red flags.
Provide a safe space for employees to report suspicious behaviour.
Conduct enhanced due diligence on suppliers, partners and customers in high-risk jurisdictions or industries.
Run ‘what-if’ scenarios for common risks and ensure frontline staff are well-equipped to spot red flags.
Review transaction data en masse to identify outliers, behavioural anomalies and unexpected patterns.
Companies lose more than 5% of their revenue each year to fraud, according to data from the Association of Certified Fraud Examiners (ACFE). The tactics deployed by fraudsters are often surprisingly mundane. They are subtle, sometimes hidden among routine payments or ordinary payroll processes and easily mistaken for human error.
Financial fraud therefore can go undetected for many months. Just last week, a former KPMG consultant was charged with siphoning nearly $1m (£700,000) through fraudulent tax returns using unauthorised credentials. He had lodged false claims for years before being detected.
Keith Williamson, managing director of the disputes and investigations team at Alvarez & Marsal, a consultancy, has worked as a forensic accountant for more than 25 years. He says the sheer volume of transactions makes fraud difficult to detect. “The number of transactions recorded by companies is phenomenal,” he says. “Compared to the past, when records were handwritten in ledgers, it’s much easier now for illicit activities to blend in. It’s also become much easier to forge documentation and data.”
