IAMI CMFI common mistakes
IAMI CMFI · Understand the traps and how to avoid them.
Treating marine fraud as a single generic category and answering questions about cargo, hull, and liability claims with the same fraud indicators, instead of distinguishing which fraud typology applies to the specific policy and claim type described.
Candidates study fraud concepts broadly and never build a mental map linking each marine line (cargo, H&M, liability, piracy) to its characteristic fraud schemes and evidence trails, so they default to whichever indicator they remember best.
What to do instead
Build a one-page matrix before the exam: rows for cargo, H&M, liability/personal injury, and theft/piracy; columns for typical fraud scheme, who usually perpetrates it, which documents reveal it, and which control prevents it. Rehearse classifying a short scenario into a row within seconds, and practise writing two or three line-specific indicators for each row.
Confusing a fraud red flag with proof of fraud, and answering that a claim should be denied or reported as fraudulent on the strength of a single indicator such as a late notification or a damaged packing list.
Red-flag lists are easy to memorise, so candidates over-apply them and forget that indicators only justify further enquiry, not a finding; exam items often test the correct next step rather than the conclusion.
What to do instead
For every indicator you learn, write the investigative step it triggers and the corroborating evidence needed before any fraud conclusion. Practise phrasing answers as 'escalate to investigation and verify X, Y, Z' rather than 'deny the claim', and check whether the question asks for a suspicion, a next action, or a determination.
Accepting cargo documents at face value, for example treating a clean bill of lading, a shipper-issued packing list, or a surveyor's report as independent confirmation of quantity, condition, or shipment date.
Candidates learn document names without learning who creates each one, who can alter it, and which documents are self-interested versus third-party; they then miss questions about staged shipments, ante-dated bills of lading, and paper-only cargo.
What to do instead
Map each cargo document to its issuer and independence level: bill of lading (carrier), commercial invoice (seller), packing list (shipper), survey report (surveyor), terminal records and vessel AIS/voyage data (third party). Drill questions by asking 'who benefits if this document is false, and what independent record would contradict it?'
Missing the physical and operational implausibilities in hull and machinery claims, such as damage patterns inconsistent with the reported machinery failure, repairs invoiced before the casualty date, or yard invoices lacking itemised labour and parts.
H&M fraud questions are heavy on technical detail and candidates without engineering or surveying background skip the plausibility check, focusing only on paperwork completeness instead of whether the damage mechanism matches the narrative.
What to do instead
Study a short list of common damage mechanisms (grounding, fatigue cracking, engine room fire, flooding, collision) and the damage signatures each produces. Practise checking a claim narrative against three anchors: timeline of the casualty, physical damage pattern, and repair invoice structure, and flag any mismatch as the answer driver.
Treating piracy, hijacking, and cargo theft as straightforward crime losses that are automatically covered, without testing for staged incidents, collusion, or misdeclared routes and cargo.
The dramatic nature of these events makes candidates accept the narrative, and the coverage questions (war risk versus cargo versus H&M, ransom and sue-and-labour issues) are complex, so they answer on emotion rather than on evidence and policy wording.
What to do instead
For each theft or piracy scenario, run a collusion and staging checklist: crew and contractor background, route deviation versus declared voyage, timing relative to loading, value and packaging of the stolen cargo versus the whole consignment, and consistency of crew accounts. Pair this with the relevant coverage clause so you can answer both the fraud and the coverage angle.
Approaching liability and personal injury claims as purely medical or quantum questions and failing to test causation, timing, and the link between the alleged injury and the insured operation.
Candidates focus on injury severity and compensation figures because those dominate the file, and they overlook the fraud dimension: pre-existing conditions, injuries occurring outside the insured period or vessel, and claims from persons not genuinely engaged in the operation.
What to do instead
Build a causation timeline for every liability scenario: date of incident, date of first medical record, date of claim notification, employment or visitor status, and witness consistency. Learn to spot the gap between the injury narrative and the insured activity, and practise identifying which record (medical, crew list, gangway log, payroll) resolves each gap.
Writing investigation plans that jump straight to interviews and document requests without defining scope, objectives, sequence, or evidence-preservation steps, and interviewing witnesses before securing physical and documentary evidence.
Candidates know the components of an investigation but not their order and dependency; under time pressure they list activities rather than design a plan, and they underestimate how early interviews allow accounts to be aligned or evidence to disappear.
What to do instead
Memorise a standard sequence: define scope and allegations, preserve and secure evidence (scene, documents, digital records, AIS and CCTV), obtain independent records, analyse, then interview. Practise drafting a five-to-eight step plan for a short scenario, and always state why each step precedes the next, especially evidence preservation before interviews.
Conducting or describing interviews in a way that leads the witness, reveals all known evidence up front, or treats an informal chat as an admissible statement, without addressing cautioning, representation, or record-keeping.
Candidates rely on general communication instincts and television-style interrogation models, and they have not studied the difference between fact-finding interviews and formal statements, or the jurisdictional rules on voluntariness and legal representation.
What to do instead
Learn a structured interview method: preparation and evidence review, non-leading open questions first, topic-by-topic probing, controlled disclosure of inconsistencies, and a contemporaneous signed record. Revise the legal limits in your jurisdiction on cautioning, right to representation, and use of statements, and practise converting a leading question into a neutral one.
Ignoring confidentiality, data protection, conflicts of interest, and jurisdictional differences when describing how evidence is obtained and shared, for example proposing to pull personal data or medical records without a lawful basis.
Candidates treat legal and ethical content as background reading and concentrate on detection techniques, so they miss questions where the technically effective action is unlawful or unethical and therefore the wrong answer.
What to do instead
For each evidence type (medical records, payroll, phone data, CCTV, AIS), note the lawful basis, consent or authorisation needed, retention limit, and who may receive it. Revise the ethical duties around conflicts, whistleblowing, and fair treatment of claimants, and practise identifying the answer that achieves the objective within the legal boundary rather than the fastest route.
Treating reporting, recovery, and prevention as an afterthought, giving vague answers such as 'report to management' or 'improve controls' without naming the report audience and content, the recovery route, or the measurable control.
Candidates concentrate on detection and investigation because those feel like the core of fraud work, and they lack familiarity with the operational outputs: suspicious activity reporting, subrogation and salvage recovery, and control design with monitoring metrics.
What to do instead
Prepare templates: a fraud report structure (findings, evidence index, quantum, recommendations, distribution list), a recovery checklist (salvage, subrogation, third-party recovery, bond or guarantee), and a control set with named metrics (pre-shipment verification rates, surveyor rotation, invoice matching exceptions, red-flag closure times). Answer with the specific artefact or metric rather than the general intention.