Where Was Internal Audit When the Fraud Happened?

By Mathuraiveran “Madhu” Marimuthu, Lead, Collaboration & Partnership at BAC Education Group and a governance, risk and internal control specialist

Six shifts internal audit must make to keep pace with modern fraud

During more than 35 years in internal audit, I have witnessed the profession evolve from transaction checking and compliance verification to risk-based auditing, governance assurance, data analytics and strategic advisory work.

Yet one expectation has remained remarkably persistent.

Whenever fraud occurs, someone will eventually ask:

“Where was internal audit?”

The question may be understandable, but it can also reveal confusion about responsibility.

Internal audit does not own the organisation’s controls, manage the fraud risk programme or guarantee that fraud will never occur. Fraud risk is a shared responsibility beginning with the board and senior management and extending across the organisation.

Nevertheless, internal audit has an indispensable role.

It can evaluate whether fraud risks are understood, whether governance arrangements are effective, whether controls are appropriately designed and operating, and whether warning signals reach the people capable of acting on them.

The IIA’s Internal Auditing and Fraud Global Practice Guide, updated in 2024 to align with the Global Internal Audit Standards, describes fraud risk as a shared organisational responsibility and provides guidance on organisational-level fraud risk assessment.

The profession must now translate that principle into a more future-ready assurance approach.

Fraud has become faster, more digital, more interconnected and increasingly supported by artificial intelligence. Internal audit cannot respond adequately by applying yesterday’s audit methods more frequently.

It must change how it sees the organisation. Six shifts can help it do so.

Shift 1: From Periodic Auditing to Continuous Risk Visibility
Traditional internal audits examine a defined period, select samples, test controls and report observations.

That model remains relevant, but it is insufficient for fraud risks that develop between audit cycles.

The ACFE’s Occupational Fraud 2026: A Report to the Nations examined 2,402 cases across 143 countries and territories. A typical fraud continued for 12 months before detection. The cases produced more than USD3.4 billion in total losses, and 20% involved losses exceeding USD1 million.

Duration matters.

A fraud that is identified in its first month is fundamentally different from one allowed to continue for a year.

Internal audit functions should identify the fraud risks that require more frequent visibility and determine whether management has appropriate monitoring in place.

Relevant indicators may include:

  • Vendor bank-account changes.
  • Duplicate addresses or contact details shared by employees and vendors.
  • Transactions just below approval thresholds.
  • Repeated emergency purchases.
  • Weekend or after-hours system activity.
  • Unusual journal entries.
  • Dormant accounts becoming active.
  • Rapid changes in customer or supplier behaviour.
  • Overrides without documented justification.

The objective is not for internal audit to become a permanent operational monitoring function.

Its role is to assess whether management has established effective monitoring, whether alerts are investigated, whether exceptions are documented and whether recurring patterns are escalated.

Where appropriate, internal audit may use continuous auditing techniques to provide assurance over high-risk areas. But responsibility for continuous monitoring should remain with management.

Shift 2: From Control Testing to Governance Assurance
Fraud does not necessarily succeed because controls are absent.

It frequently succeeds because the people with authority can override them, because departments fail to share information or because employees believe challenging senior management is unsafe.

Internal audit must therefore look beyond the control activity itself.

Suppose an organisation requires three quotations for procurement. An audit could confirm that three documents were present and conclude that the control operated.

A governance-centred audit would ask additional questions:

  • Were the bidders genuinely independent?
  • Do they share directors, addresses or contact details?
  • Was the specification designed for a preferred vendor?
  • Who selected the vendors invited to bid?
  • Were unsuccessful quotations authentic?
  • Did the approver have an undisclosed relationship?
  • Were exceptions concentrated around one officer?
  • Did employees feel able to challenge the process?

This is the difference between verifying compliance and assessing whether the control achieved its intended purpose.

The Global Internal Audit Standards, effective from January 9, 2025, reinforce internal audit’s purpose in strengthening governance, risk management and control. They are structured across five domains covering the purpose of internal auditing, ethics and professionalism, governance of the function, management of the function and performance of internal audit services.

Fraud assurance should reflect that wider purpose.

Internal audit must examine tone at the top, decision-making authority, management override, board information flows and organisational culture—not merely procedural compliance.

Shift 3: From Transactional Red Flags to Behavioural and Cultural Signals
Data can show what happened.

Behaviour often helps explain why.

The ACFE’s 2026 report found that 84% of fraudsters displayed at least one behavioural red flag. The indicators associated with particularly high median losses included excessive internal pressure, past legal problems, unusually close relationships with vendors or customers, refusal to take leave and a “wheeler-dealer” attitude.

Internal auditors should not treat behavioural indicators as proof of fraud.

A person working long hours or declining leave may be entirely innocent. Behaviour becomes relevant when combined with control weaknesses, unusual transactions, complaints, conflicts of interest or signs that an employee is sharing confidential information with vendors.

The larger issue is culture.

Do employees believe performance targets must be achieved at any cost? Are people rewarded for results without regard to how those results were produced? Are executives exempted from normal controls? Does bad news travel upward? Are previous misconduct cases resolved consistently?

The IIA’s updated guidance on auditing culture recognises that culture affects the control environment and provides internal auditors with approaches for assessing associated risks.

Fraud risk assessment should therefore include cultural drivers such as pressure, fear, excessive loyalty, tolerance of exceptions and concentration of power.

Shift 4: From Organisational Boundaries to Ecosystem Risk
Many fraud risks now originate outside the traditional organisational perimeter.

Businesses depend on vendors, agents, contractors, consultants, outsourced service providers, digital platforms, joint ventures and supply-chain partners.

These relationships create value, but they also create distance between the organisation and the conduct performed on its behalf.

A vendor may collude with an employee. An intermediary may make an improper payment. An outsourced provider may manipulate service data. A technology partner may have excessive access to sensitive systems.

Internal audit plans should therefore consider the complete third-party lifecycle:

  • Selection and onboarding.
  • Beneficial ownership.
  • Conflict-of-interest checks.
  • Due diligence.
  • Contractual safeguards.
  • System access.
  • Performance verification.
  • Payment controls.
  • Changes in ownership or bank details.
  • Ongoing monitoring.
  • Renewal and termination.

A strong contract is not a substitute for monitoring.

Nor should due diligence be treated as a one-time onboarding activity. Risk changes as ownership, personnel, financial condition and political exposure change.

Internal auditors should assess whether third-party governance is proportionate to risk and whether information is shared across procurement, legal, compliance, finance, operations and cybersecurity.

Fraud frequently flourishes in the spaces between these functions.

For Malaysian organisations, this is a legal question as well as a control question. Section 17A of the Malaysian Anti-Corruption Commission Act 2009, in force since June 2020, makes a commercial organisation liable for corruption committed by persons associated with it, including those performing services on its behalf. Its principal defence is to show that adequate procedures were in place to prevent such conduct.

Internal audit is well placed to tell the board whether those procedures work in practice across the third-party lifecycle, or exist mainly on paper.

Shift 5: From Investigating Incidents to Learning From Them
Organisations often treat fraud investigations as isolated and confidential events.

Confidentiality is necessary, but excessive secrecy can prevent organisational learning.

Once an investigation is complete, the organisation should ask:

  • Which control failed?
  • Was the control bypassed, poorly designed or never performed?
  • Were warning signs previously identified?
  • Did anyone raise a concern?
  • How was the concern handled?
  • Did incentives contribute to the misconduct?
  • Were similar transactions reviewed?
  • Was recovery pursued?
  • Were disciplinary decisions consistent?
  • Were findings shared with the board?
  • Have corrective actions been independently validated?

Internal audit may conduct investigations where permitted by its mandate and where the necessary competence and independence exist. It may also support or review investigations led by other functions.

However, internal audit should not automatically become the organisation’s default investigation owner.

The chief audit executive must consider whether investigative responsibilities could compromise—or appear to compromise—the function’s ability to provide subsequent independent assurance.

Where internal audit has participated directly, appropriate safeguards may include external review, different personnel, clear reporting protocols and disclosure of potential impairments.

The greater contribution of internal audit may come after the case: assessing whether management identified the root causes and whether remediation addresses the organisational conditions that allowed the fraud to occur.

Shift 6: From Using AI to Auditing AI
Internal audit functions are increasingly using AI to review documents, analyse transactions, identify anomalies, summarise information and assist with audit planning.

These tools can improve coverage and productivity.

But internal auditors must resist automation bias—the tendency to accept a technological output because it appears objective or sophisticated.

An AI-generated alert may be incorrect. A model may reflect incomplete, biased or manipulated data. A fraudster who understands the detection rules may structure transactions to avoid them.

Internal audit needs governance around its own use of AI:

  • Which tools are approved?
  • What data may be uploaded?
  • How are confidential records protected?
  • Are outputs independently validated?
  • Are prompts and methodologies documented?
  • Can the result be explained?
  • Who remains accountable for the conclusion?
  • Are legal, privacy and ethical requirements addressed?

The same questions should be asked when auditing management’s use of AI.

NIST’s AI Risk Management Framework and its Generative AI Profile emphasise risk management across the AI lifecycle, including design, deployment, use, testing and evaluation.

AI governance should become part of the fraud assurance universe because poorly governed AI can create new opportunities for manipulation, unauthorised decisions, data leakage and concealed bias.

Whistleblowing Remains Indispensable

The most technologically advanced fraud programme can still fail if employees are afraid to speak.

In the ACFE’s 2026 study, 43% of frauds were detected following tips, and more than half of those tips came from employees.

In Malaysia, statutory protection under the Whistleblower Protection Act 2010, amended in 2025, centres on disclosures made to enforcement agencies. Whether employees use internal channels therefore depends heavily on whether they trust them.

Internal audit should evaluate more than the existence of a hotline.

It should consider:

  • Whether channels are accessible to employees and external parties.
  • Whether anonymous reporting is available where legally permitted.
  • Who receives and triages allegations.
  • Whether conflicts are identified.
  • How retaliation risks are assessed.
  • Whether allegations involving senior management bypass management.
  • Whether investigations are timely.
  • Whether cases are closed with adequate evidence.
  • Whether trends are reported to the board.
  • Whether employees trust the system.

A hotline without trust is only a communication channel.

Internal audit can provide assurance over the system, but management and the board must create the culture that makes people willing to use it.

The CAE’s Boardroom Conversation

Chief audit executives should move the fraud discussion beyond the number of reported cases.

A low number of allegations may indicate low fraud exposure. It may also indicate fear, weak awareness or ineffective reporting channels.

The board needs a more complete picture.

In Malaysia, the Malaysian Code on Corporate Governance expects boards to take ownership of anti-corruption and whistleblowing arrangements, not merely to receive reports about them.

Useful reporting may include:

  • The organisation’s principal fraud risks.
  • Areas of high residual exposure.
  • Significant overrides and exceptions.
  • Whistleblowing trends.
  • Ageing investigations.
  • Losses, recoveries and disciplinary outcomes.
  • Third-party concerns.
  • Repeat incidents.
  • Remediation status.
  • Emerging technology-enabled threats.
  • Gaps in fraud awareness and capability.

The most important board question may be:

“What are we not currently seeing?”

A Profession Built on Courage

When I reflect on my years serving the internal audit profession, including my tenure as President and long-standing board member of IIA Malaysia, I am reminded that the profession’s value does not come from producing more reports.

It comes from helping organisations confront reality.

Internal auditors must be technologically capable, but they must also remain curious, independent and willing to ask uncomfortable questions.

We should use AI, but not surrender judgment to it.

We should analyse data, but not ignore human behaviour.

We should understand management’s pressures, but not allow empathy to weaken objectivity.

We should support organisational success, but remember that our greatest responsibility, in the words of the Global Internal Audit Standards, is to help the organisation create, protect and sustain value through independent, risk-based and objective assurance, advice, insight and foresight.

Fraud will continue to evolve.

The enduring defence will be a profession prepared to evolve with it—without compromising the integrity on which its credibility depends.

So when the next fraud surfaces and someone asks, “Where was internal audit?”, the best possible answer is this: already asking the questions no one else wanted to ask.

About the Writer
Mathuraiveran “Madhu” Marimuthu is Lead, Collaboration & Partnership at BAC Education Group and a governance, risk and internal control specialist with more than 35 years of internal audit, fraud risk management, corporate governance and board oversight experience. A Chartered Accountant and Certified Internal Auditor (CIA), he is a Past President of the Institute of Internal Auditors Malaysia and served for approximately 20 years on the institute’s board. He brought the CIA certification to Malaysia in 1994. He has served on corporate boards and Audit Committees, is an HRD Corp Accredited Trainer, and is the author of the Fraud Detection & Prevention Toolkit (BAC Education, 2026).

To learn more about BAC Education Group’s internal audit and fraud risk programmes, including CIA certification preparation, visit www.HRDAcademy.asia.

References

  • Association of Certified Fraud Examiners, Occupational Fraud 2026: A Report to the Nations.
  • The Institute of Internal Auditors, Global Internal Audit Standards (effective 9 January 2025).
  • The Institute of Internal Auditors, Global Practice Guide: Internal Auditing and Fraud, 3rd Edition (2024).
  • The Institute of Internal Auditors, Global Practice Guide: Auditing Culture, 2nd Edition (2024).
  • National Institute of Standards and Technology, AI Risk Management Framework and Generative AI Profile.
  • Malaysian Anti-Corruption Commission Act 2009, Section 17A.
  • Whistleblower Protection Act 2010 (as amended in 2025).
  • Malaysian Code on Corporate Governance.

More for You