1What is adverse media screening?

Searching news and other public sources for negative information about a customer, beneficial owner or counterparty — allegations, investigations, charges, convictions, regulatory action or reputationally damaging conduct. It supplements list-based screening with information no list contains.

2Why is adverse media screening required?

Because sanctions and PEP lists only capture designated or defined categories. Someone under criminal investigation for fraud appears on no list, yet clearly presents risk. Adverse media closes that gap and is expected as part of customer due diligence under FATF-aligned regimes.

3How does it differ from sanctions screening?

Sanctions screening is a binary match against an official list with a legal consequence. Adverse media is unstructured, judgement-based, and produces risk information rather than a prohibition. There is no equivalent of a "true match" that forces a specific action.

4Is adverse media the same as negative news?

The terms are used interchangeably. "Adverse media" is the regulatory phrasing; "negative news" is the operational shorthand. Both mean publicly available information suggesting risk.

5When should adverse media screening take place?

At onboarding, at periodic review, on trigger events such as a change in ownership or unusual activity, and on an ongoing basis for higher-risk customers. Risk emerges after onboarding, so a one-off check is insufficient.

6Who should be screened?

The customer, beneficial owners, directors and controllers, and in higher-risk cases key counterparties and connected parties. Screening only the named account holder misses exposure sitting behind the structure.

7Does adverse media apply to all customers equally?

No — it should be risk-based. Standard-risk retail customers may receive automated screening only; higher-risk customers, PEPs and complex corporates warrant deeper, manually reviewed searches.

8What outcome can adverse media lead to?

A change in risk rating, enhanced due diligence, senior escalation, a suspicious activity report, or exit. It can also lead to no action at all where the information is irrelevant or not credible — that is a legitimate documented outcome.

9Does adverse media alone justify exiting a customer?

Rarely on its own. It should inform a risk assessment rather than trigger automatic exit. Reacting to unverified reporting risks unfair exclusion and is a form of de-risking regulators criticise.

10How does adverse media relate to reputational risk?

They overlap but are not identical. Financial crime risk asks whether the customer may be involved in illicit activity; reputational risk asks whether association damages the firm. Something can be reputationally awkward without being a financial crime concern, and vice versa.

11What makes adverse media harder than other screening?

The data is unstructured, sources vary wildly in reliability, names are ambiguous, coverage differs by language and jurisdiction, and there is no authoritative answer to check against. Judgement carries the weight.

12What is the cost of getting it wrong in each direction?

Over-reacting excludes legitimate customers and invites de-risking criticism. Under-reacting means onboarding someone under active investigation, which is exactly the failure enforcement actions highlight. Both are real, which is why documentation matters more than the outcome itself.

13What sources are used for adverse media screening?

Mainstream and financial news, regulatory and enforcement notices, court records and litigation databases, law enforcement announcements, corporate registries, sanctions and debarment lists, and investigative journalism outlets.

14What is a structured versus unstructured source?

Structured sources are databases with defined fields — court records, regulatory registers, debarment lists. Unstructured sources are free text such as news articles. Structured data is easier to match reliably; unstructured data carries most of the useful nuance.

15Why are official sources weighted more heavily?

Because regulatory notices, court filings and enforcement announcements are verifiable and carry legal standing. A regulator's published fine is fact; a newspaper allegation is a report of a claim.

16How do commercial adverse media tools work?

They aggregate large volumes of media and public records, apply entity resolution to link articles to individuals or companies, and categorise content by risk type. The firm still reviews the output — the tool narrows the field, it does not decide.

17What are the limitations of commercial tools?

Language and regional coverage gaps, latency in indexing, weak entity resolution producing wrong-person matches, over-broad categorisation, and reliance on sources that vary in reliability. Coverage is never complete.

18Is a plain internet search sufficient?

Not for a compliance-grade check. Search results are personalised, non-reproducible, lack archives, and miss non-indexed sources. It is useful as a supplement, but the audit trail requires a systematic tool or documented method.

19Why does language coverage matter?

Because significant reporting often exists only in local-language media. A customer with no adverse media in English may have extensive coverage in their home market, so screening must extend beyond English sources for international exposure.

20How do you handle transliteration in searches?

Search known variants of the name, use native-script searches where possible, and check for common alternative spellings. Missing a variant is one of the most frequent causes of a false negative.

21What is an investigative journalism consortium and why does it matter?

Cross-border collaborations that publish large document-based investigations into offshore structures and financial crime. Their releases frequently trigger mass re-screening exercises because they name previously unreported parties.

22How reliable is social media as a source?

Generally weak on its own — unverified, easily fabricated and often anonymous. It can indicate a lead worth checking against credible sources, but it should not by itself drive a risk decision.

23What are debarment lists?

Registers of parties excluded from contracting with governments or multilateral institutions, usually for fraud or corruption. They are highly relevant because exclusion follows a formal finding rather than an allegation.

24Should you record the source of every finding?

Always. Capture the publication, date, URL and a summary. Articles are removed or paywalled over time, so a file that references "press reports" without specifics cannot be reviewed or defended later.

25How do you handle paywalled or removed articles?

Record the citation details and a factual summary at the time of review, and note access limitations. Where permitted, retain an archived copy consistent with the firm's data retention and copyright policies.

26How far back should searches go?

Risk-based. Recent material matters most, but serious historic matters — a fraud conviction, a major enforcement action — remain relevant for years. The firm's policy should set the default lookback period.

27What is entity resolution?

Determining whether the person or company in an article is the same as your customer. It is the core technical challenge in adverse media, because names alone rarely identify anyone uniquely.

28What identifiers help resolve a match?

Date of birth, nationality, place of residence, occupation, employer, company registration number and known associates. Occupation and location resolve most consumer-level ambiguity quickly.

29Why do common names create such difficulty?

Because a name shared by thousands generates constant matches with no discriminating detail. In some jurisdictions naming conventions compound this, making occupation and location the only practical differentiators.

30How would you tune adverse media screening?

Segment by customer risk so higher-risk relationships get broader searching, refine risk categories so irrelevant content is filtered, improve customer identifier quality, and review outcomes to see which settings produce genuine escalations.

31What is category filtering?

Restricting results to defined risk types — financial crime, corruption, terrorism, organised crime, regulatory action — rather than all negative coverage. It reduces noise but risks filtering out relevant material, so categories must be set deliberately.

32Should all negative news be captured, or only financial crime news?

It depends on the firm's risk appetite and policy. Most focus on financial crime, corruption and regulatory categories, but some include violence, environmental offences or human rights where reputational exposure matters.

33How do you avoid alert fatigue?

Better filtering, better customer data, whitelisting resolved false positives with periodic revalidation, and routing genuinely low-risk customers to lighter screening. Reducing volume by loosening categories without analysis just hides risk.

34What is a whitelist in adverse media, and what is the danger?

A record of previously discounted matches so the same wrong-person result is not re-investigated repeatedly. The danger is staleness — a match discounted correctly last year may relate to genuinely new reporting now, so entries need periodic review.

35How frequently should ongoing screening run?

Risk-based — continuous or daily monitoring for high-risk customers and PEPs, periodic for standard risk. Continuous monitoring is increasingly the norm for higher-risk portfolios because it catches events between reviews.

36What is event-driven review?

Triggering a review when new adverse media appears, rather than waiting for the scheduled cycle. It is more responsive and reduces the window in which a firm is unaware of emerging risk.

37How do you screen a corporate customer?

Screen the entity, its beneficial owners, directors and senior management, and trading names or former names. Corporate risk frequently attaches to the individuals behind it rather than the registered company.

38How do you handle a company that has changed its name?

Screen both current and former names. Rebranding after adverse publicity is common, and searching only the current name will return nothing.

39How would you test whether adverse media screening works?

Inject known cases — individuals with well-documented public findings — and confirm they are returned. Also sample high-risk customers manually to see whether anything material was missed. Alert volume alone tells you nothing about coverage.

40What role does AI play in adverse media screening?

It improves entity resolution, classifies articles by risk type, summarises long coverage and reduces duplication. It does not remove the need for human judgement on credibility and materiality, and its decisions still need to be explainable.

41Walk me through assessing an adverse media hit.

First, is it the same person or entity — compare identifiers. Second, is the source credible and independent. Third, what is the nature of the matter and its stage: allegation, investigation, charge, conviction, or dismissal. Fourth, is it material to financial crime risk and how recent. Then I document the finding, the reasoning and the recommended action.

42What is the difference between an allegation and a conviction?

An allegation is an unproven claim; a charge means formal proceedings have begun; a conviction is a judicial finding. They carry escalating weight, and treating an allegation as equivalent to a conviction is both unfair and analytically wrong.

43How do you assess source credibility?

Consider the outlet's reputation and editorial standards, independence from state or commercial interests, whether the reporting is corroborated elsewhere, whether it names sources or documents, and whether it has been retracted or challenged.

44What weight would you give a single uncorroborated source?

Limited on its own, but not zero. I would search for corroboration and, absent it, record the finding with a note on its unverified status and factor it into the risk rating rather than treating it as established.

45How do you handle adverse media in a jurisdiction with a controlled press?

Cautiously in both directions. Reporting may be politically motivated; absence of reporting may reflect suppression rather than clean conduct. I would weight source independence heavily and look for international corroboration.

46What if the customer was acquitted or the case was dropped?

Record the full outcome. An acquittal or discontinuance substantially reduces the weight of the matter, though the underlying circumstances may still be relevant to context. It should never be recorded as though the allegation still stands.

47How does the age of a matter affect its weight?

Recent matters carry more weight, but seriousness matters more than recency. A twenty-year-old parking dispute is irrelevant; a twenty-year-old fraud conviction may still be material, particularly for a senior role or high-risk relationship.

48What is materiality in this context?

Whether the information actually bears on financial crime or reputational risk for this relationship. A civil contract dispute is negative coverage but usually immaterial; a money laundering investigation clearly is.

49Would you escalate every confirmed adverse media finding?

No. Confirmed but immaterial findings are documented and closed. Escalation should be driven by materiality and risk, not by the mere existence of a true match — otherwise senior reviewers are flooded with noise.

50What if the customer disputes the reporting?

Take the response seriously and record it. Ask for supporting documentation — a judgment, a retraction, a regulatory outcome. A customer who can evidence that reporting was wrong should not carry the risk of it indefinitely.

51Can you ask the customer directly about adverse media?

Usually yes, and it is often the fastest way to resolve ambiguity. The exception is where doing so would tip off in relation to a suspicion or report — in that case the enquiry must not proceed without guidance.

52What documentation should support your conclusion?

The search performed and sources covered, the results found with citations and dates, the identity assessment, the credibility and materiality analysis, the conclusion and the action taken. Another reviewer should be able to follow it without repeating the search.

53What is the most common documentation weakness?

Recording a conclusion without the reasoning — "no material adverse media found" with no record of what was searched or reviewed. That cannot be tested by audit and is a frequent regulatory criticism.

54How do you avoid bias in assessment?

Work from documented criteria rather than instinct, be conscious that prominence or nationality is not evidence, apply the same standard regardless of the customer's commercial importance, and have higher-impact decisions independently reviewed.

55What if adverse media relates to a company the customer left years ago?

Assess their role and tenure. A director during the conduct in question is materially different from an employee who joined afterwards. Timeline matters as much as association.

56How does adverse media interact with a suspicious activity report?

It can support suspicion but is rarely sufficient alone. Where credible adverse media aligns with unexplained activity in the account, the combination is often what converts concern into a reportable suspicion.

57What risk categories does adverse media typically cover?

Money laundering, fraud, corruption and bribery, terrorism and terrorist financing, organised crime, sanctions evasion, tax offences, market abuse, regulatory enforcement, and in some frameworks trafficking, environmental crime and human rights abuses.

58Which categories carry most weight?

Predicate offences directly linked to financial crime — money laundering, fraud, corruption, sanctions evasion, organised crime. These bear most directly on whether the customer's funds may be illicit.

59Why does regulatory enforcement matter even without criminal proceedings?

Because it reflects a formal finding by an authority following investigation. A regulatory fine for control failures is verifiable fact, which gives it far more weight than an unproven media claim.

60How does adverse media support PEP due diligence?

PEP status identifies the role; adverse media indicates whether concerns attach to the individual. It is also central to testing source of wealth — corruption allegations directly undermine a claimed legitimate origin.

61What adverse media would concern you most for a corporate customer?

Investigations into the company or its directors, regulatory enforcement, debarment from public contracting, insolvency or fraud proceedings, and reporting linking it to sanctioned or criminal parties.

62What is predicate offence reporting and why does it matter?

Coverage of the underlying crime that generates illicit funds — fraud, drug trafficking, corruption. It matters because laundering follows a predicate offence, so evidence of one is direct evidence of laundering risk.

63How would you treat adverse media about tax matters?

Distinguish avoidance from evasion. Aggressive but lawful planning is not a financial crime concern in most regimes; evasion is a predicate offence. Media coverage frequently conflates the two, so the underlying facts matter.

64What about environmental or human rights reporting?

Increasingly in scope, particularly for corporate customers and where ESG considerations feed risk appetite. Whether it drives a financial crime decision depends on the firm's stated policy.

65What is reputational contagion?

The risk that association with a customer damages the firm even where no legal breach occurs. It is a legitimate factor in relationship decisions but should be assessed separately from financial crime risk, not merged with it.

66How does adverse media help detect networks?

Reporting often names multiple connected individuals and entities. Cross-referencing those names against the customer base can reveal relationships that structured data does not show.

67What is a leak-driven review?

A mass re-screening exercise following publication of a major document leak, where firms check their customer base against newly published names. They are resource-intensive and time-critical.

68How would you handle a customer named in a large leak?

Being named is not itself wrongdoing — many named parties have entirely lawful arrangements. I would establish what the documents actually show about their role, assess it against known information such as source of wealth, and escalate only where there is genuine inconsistency.

69What if adverse media concerns a customer's counterparty rather than the customer?

It still matters. Exposure to a counterparty under investigation can taint funds and indicate the customer's own risk. I would assess the nature and frequency of dealings before drawing conclusions.

70How does adverse media inform customer risk rating?

Credible, material findings raise the rating, which in turn drives enhanced due diligence, tighter monitoring and shorter review cycles. The rating change is the mechanism through which the finding actually affects controls.

71What should an adverse media policy define?

Scope of screening and who is covered, risk categories in scope, sources and tools used, frequency by risk tier, credibility and materiality criteria, escalation routes, documentation standards and retention.

72Why does the policy need explicit credibility criteria?

Because without them, assessments vary by analyst and cannot be defended as consistent. Written criteria turn individual judgement into a repeatable, auditable standard.

73Who decides on escalation?

Analysts assess and recommend; escalation thresholds are set in policy; material decisions sit with compliance management or a committee. Where exit is contemplated, the decision belongs at senior level.

74What management information would you report?

Screening volumes and hit rates, confirmed versus discounted matches, escalations by category, resulting risk rating changes and exits, ageing of unreviewed alerts, and coverage or tool performance issues.

75What would audit test?

Whether screening scope matches policy, whether searches were actually performed and documented, whether identity and credibility assessments are reasoned, whether escalation thresholds were applied consistently, and whether findings changed anything.

76What are the most common programme failures?

Screening only the account holder, English-only searching, conclusions without recorded reasoning, stale whitelists, treating all negative coverage as equivalent, and findings that never result in any change to risk rating or monitoring.

77How do data protection rules affect adverse media?

Screening processes personal data, generally under a legal obligation basis. Accuracy obligations are significant — recording someone as linked to crime on weak evidence creates both regulatory and legal exposure.

78What if you record a finding and it turns out to be the wrong person?

Correct the record promptly and document the correction and its basis. Leaving an erroneous association in a customer file is a data accuracy failure with real consequences for that individual.

79How long should adverse media findings be retained?

In line with the firm's AML retention policy, commonly five years from the end of the relationship. Findings should also be revisited at review rather than left to sit indefinitely unexamined.

80How should adverse media training be structured?

Focused on judgement — how to assess sources, distinguish stages of legal process, judge materiality and document reasoning. Case-based training is far more effective than definitional training for this control.

81How do you ensure consistency across a team?

Written criteria, worked examples, calibration sessions where analysts assess the same case and compare, quality assurance sampling, and a clear escalation path for borderline judgements.

82Should adverse media screening be outsourced?

Search and collation can be, but the risk decision cannot be delegated away — accountability stays with the firm. Outsourced arrangements need oversight, quality testing and clear standards.

83What is the second line's role?

Setting policy and criteria, challenging assessments, monitoring MI, testing quality, and having authority to override a business decision to retain a relationship.

84How would you demonstrate the programme is effective?

Through outcomes, not volumes — findings that changed risk ratings, triggered EDD or led to reports; coverage testing showing known cases are detected; and QA showing consistent, well-reasoned decisions.

85A customer shares a name with someone convicted of fraud abroad. There is no date of birth in the article. What do you do?

I would look for other discriminators — nationality, occupation, location, employer, age indicators, associated companies. If the convicted person operated in a different country and industry, that supports a documented discount. If nothing distinguishes them, I would treat it as unresolved and escalate rather than assume it is not the same person.

86Adverse media reports your customer is "under investigation" but names no authority and cites no source. How much weight?

Little on its own. I would search for corroboration from official sources or credible outlets. Absent that, I would record it as an unverified single-source claim, note it in the risk assessment, and monitor for further reporting rather than act on it.

87A corporate customer's former CEO was convicted of bribery. The CEO left three years before the conduct. Your assessment?

Timeline is decisive. If the conduct occurred after their departure, the relevance to the current relationship is limited. I would confirm the dates, check whether the company itself faced findings, and document why the association does not carry material risk.

88A high-value customer appears in credible reporting on a money laundering investigation. The relationship manager objects to escalation. What do you do?

Escalate regardless. Commercial value is irrelevant to whether risk exists, and the objection itself is worth documenting. I would present the evidence, follow the escalation path, and ensure the decision is taken at the appropriate level rather than settled between us.

89Your tool returns 400 articles on a common-named customer. How do you handle it?

Work systematically rather than reading all 400. Filter by risk category and jurisdiction, use identifiers to eliminate obvious non-matches in bulk, and focus on the highest-severity categories first. Then document the method used, not just the conclusion.

90Adverse media appears only in local-language sources you cannot read. What do you do?

Do not close it as unreviewable. Use translation, escalate to a colleague with the language, or engage the firm's research function. Language barriers are a resourcing issue, not a reason to record no findings.

91A customer provides a court judgment showing an earlier reported allegation was dismissed. How do you record it?

Update the file to reflect the outcome clearly, retain both the original finding and the dismissal, and adjust the risk assessment accordingly. The record should show the complete picture, not just the negative half.

92You find adverse media on a beneficial owner who is not the account holder. Does it matter?

Yes, materially. Beneficial owners control the funds and the entity, so risk attaching to them attaches to the relationship. I would assess it exactly as I would for a direct customer.

93A major leak is published and 12 of your customers are named. What is your approach?

Triage by exposure rather than treating all 12 identically — what role each played, whether structures were disclosed to us previously, and whether their known source of wealth is consistent. Being named is not wrongdoing, so I would assess individually and escalate only where there is genuine inconsistency.

94You realise a previous analyst closed a hit that you believe was material. What do you do?

Reopen and reassess rather than leave it. I would document the new assessment and reasoning, flag it through QA so any systemic issue is identified, and avoid framing it as blame — the point is correcting the risk position.

95Why does adverse media work interest you?

Give a genuine answer. Strong responses reference the research and reasoning involved, the fact that no two cases are identical, and the responsibility of making a judgement call that affects both the firm and the customer.

96How do you handle ambiguity when there is no clear right answer?

Apply the documented criteria, reason transparently, take the more cautious route where genuinely balanced, and escalate rather than force a conclusion. Ambiguity handled openly is defensible; a false show of certainty is not.

97Tell me about a time you had to weigh conflicting information.

Use a structured real example — the conflict, how you evaluated each side, what you decided and why. If your experience is training-based, say so; invented examples collapse under follow-up questions.

98How do you avoid letting personal opinion influence an assessment?

By anchoring to evidence and criteria, separating what is reported from what is proven, and being alert to assumptions based on nationality, industry or prominence. Independent review on significant decisions helps too.

99How do you stay current in this area?

Regulator enforcement notices, FATF guidance, investigative journalism reporting, and industry publications on screening practice. Enforcement cases are especially useful because they show which specific failures are being penalised.

100What questions do you have for us?

Ask which tool and sources the team uses, how language coverage is handled, how credibility criteria are defined, how escalation thresholds are set, and how the team managed the last major leak-driven review. These show you understand where the real difficulty lies.