Adverse Media Screening: What It Is and How It Works

Adverse media screening is the check a regulated business runs to find out whether a customer, a company's owners or a counterparty appear in negative news or other public information linked to financial crime, such as fraud, corruption, money laundering, sanctions evasion or terrorism. It sits alongside sanctions and PEP screening as part of customer due diligence, and it matters most when a firm needs to decide whether a customer is higher risk.

This guide explains what counts as adverse media, how screening works in practice, why false positives are the hard part, where AI tools fit, what EU law says, and how crypto exchanges use it. It is written for compliance teams and for users wondering why an exchange asked about a news story.

What counts as adverse media?

Adverse media (also called negative news) is any credible public information suggesting a person or company may be involved in financial crime or serious wrongdoing. Typical sources are:

• News outlets: national and local press, business and trade publications, in many languages.
• Official publications: regulator enforcement notices, court judgments and prosecutors' press releases.
• Investigative and civil society reports: research by journalists' networks and NGOs.
• Other public sources: company registries, blogs and social media, which usually carry less weight on their own.

Most screening programmes group results by theme: financial crime and fraud, bribery and corruption, organised crime, terrorism, sanctions evasion, tax crime, cybercrime and so on. A story about a parking dispute or a celebrity divorce is not adverse media in this sense. Relevance to money laundering risk is what counts.

Is adverse media screening a legal requirement?

Usually not by name. Neither the EU's Anti-Money Laundering Regulation nor the US program rules use the phrase "adverse media screening". Instead, they require firms to understand who their customers are and how risky they are, and checking public information is one of the main ways firms do that.

• EU. The Anti-Money Laundering Regulation, Regulation (EU) 2024/1624 (AMLR), applies from 10 July 2027. Its Annex I lists the risk variables firms must consider when deciding how much due diligence to apply, and the customer variables include "the customer's and the customer's beneficial owner's reputation". Recital 31 also names media among the "credible and reliable sources" that can inform a firm's risk assessment.
• US. FinCEN's program rule for banks, 31 CFR 1020.210, requires risk-based customer due diligence that builds a customer risk profile and includes ongoing monitoring. Adverse media findings are a common input to that profile.

In practice, supervisors expect to see public-information checks in enhanced due diligence files, especially for politically exposed persons, high-risk corporate clients and large or unusual transactions. Our guide to CDD vs EDD covers when those heavier checks apply.

How adverse media screening works

A typical process has four steps:

1. Search. The customer's name, plus known aliases and the names of company directors and beneficial owners, are run against news archives and risk databases. Good systems search in several languages and handle transliterated names.
2. Match. Results are matched to the person using other identifiers, such as date of birth, nationality, location or job title, to rule out namesakes.
3. Classify. Each relevant story is tagged by crime category, date, source quality and how far the case has gone: a rumour, an allegation, an investigation, charges or a conviction.
4. Decide and record. An analyst decides whether the finding changes the customer's risk rating, needs more questions, or is irrelevant, and writes down why.

Screening happens at onboarding and then again over time. Many firms re-screen their customer base automatically and alert on new stories, so a customer who was clean at sign-up can be flagged later. Vendors bundle this with sanctions and PEP checks. iDenfy's AML screening, for example, combines adverse media checks with sanctions and PEP lists and ongoing screening. Our comparison of sanctions screening software and our guide to watchlist and PEP screening cover the list-based side.

False positives and the allegation problem

Adverse media is noisier than a sanctions list. A sanctions list is an official record; a news archive is not. Two problems dominate:

• Namesakes. Common names return hundreds of irrelevant stories. Without secondary identifiers, analysts waste time and good customers get delayed.
• Allegations are not convictions. A story may report an accusation that was later dropped, or a case the person won.

EU law addresses the second point directly. Article 76 of the AMLR lets obliged entities process data on criminal offences only if the data come from reliable sources, are accurate and up to date, and the firm has procedures that distinguish between "allegations, investigations, proceedings and convictions", taking into account the right to a fair trial and the presumption of innocence. In practice, an old, unproven allegation should not weigh the same as a recent conviction, and the file should show that the analyst knew the difference.

Firms tune their screening to cut noise: matching on more identifiers, limiting results to relevant crime categories, ranking sources by credibility and suppressing stories already reviewed. The goal is not zero hits but hits that are worth an analyst's time.

AI and NLP in adverse media screening

Modern tools use natural language processing to read articles rather than just match keywords. Typical features are entity recognition (is this article actually about the customer?), topic classification (fraud, corruption, terrorism), sentiment and severity scoring, translation, and deduplication of the same story across hundreds of outlets.

These tools reduce manual review, but the AMLR sets limits on automated decisions. Under Article 76(5), firms may use automated processes or AI systems only if:

• the data used are limited to data collected under the regulation's due diligence chapter;
• any decision to accept, refuse or end a customer relationship, or to increase or decrease the level of due diligence, is subject to "meaningful human intervention"; and
• the customer can get an explanation of the decision and challenge it, except where it concerns a suspicious transaction report.

So AI can sort and summarise, but a person still has to own the decision.

Where adverse media fits in CDD and EDD

An adverse media hit rarely means an automatic rejection. More often it moves a customer into enhanced due diligence, where the firm asks for context and evidence, such as source of funds and source of wealth. If the story suggests the customer may be laundering money, the firm may also have to consider a report to its financial intelligence unit, which it must not disclose to the customer.

Adverse media screening in crypto

Crypto-asset service providers are obliged entities under the AMLR, so the same expectations apply to them. Common crypto use cases are:

• Corporate onboarding. Screening funds, trading firms and token issuers, their directors and owners, as part of KYB.
• Token listings and partnerships. Checking project founders and promoters for past fraud, rug pulls or enforcement actions before listing or partnering.
• OTC and institutional desks. Screening counterparties in large trades, where one bad actor can carry real risk.
• Other exchanges. Assessing crypto firms the exchange deals with, alongside licensing and Travel Rule checks.
• Retail EDD. Reviewing high-value customers and anyone flagged by on-chain signals.

Adverse media screening covers people. It does not see wallets. Exchanges pair it with blockchain analytics, which traces funds on-chain, and with transaction monitoring. A customer with clean news coverage can still be flagged for depositing from a mixer, and the reverse is true too. How these signals feed a firm's overall risk model is covered in our guide to AML risk assessment for crypto businesses.

For users: why an exchange asked about a news story

If an exchange asks about an article, it has usually found a possible match and needs to rule it in or out. Reply directly: if it is not you, say so and provide identifiers that show it; if it is you, explain the outcome and share documents such as a court decision showing charges were dropped. Ignoring the request is the most likely route to a restricted account.

JewelSwap's apps are non-custodial DeFi software and do not screen users themselves; our explainer on KYC in DeFi covers why.

Frequently asked questions

What is adverse media screening?

Adverse media screening is the process of checking news and other public sources for information linking a customer, its owners or a counterparty to financial crime, such as fraud, corruption, money laundering, sanctions evasion or terrorism. It is part of customer due diligence and is used to set a customer's risk rating.

Is adverse media screening required by law?

Usually not by name. EU and US rules require firms to understand customer risk, and the EU AMLR lists a customer's reputation as a risk variable. Adverse media screening is the standard way firms meet that expectation, especially in enhanced due diligence.

What is the difference between adverse media and sanctions screening?

Sanctions screening checks names against official lists, and a confirmed match generally means assets must be frozen and the firm cannot deal with the person. Adverse media screening searches news and public sources, which are less reliable, so hits need human review and usually lead to more questions rather than automatic refusal.

Does an adverse media hit mean I will be rejected?

Not necessarily. Many hits are namesakes or old allegations. A relevant hit usually means the firm asks for more information and may apply enhanced due diligence. Under EU rules, firms must distinguish between allegations, investigations, proceedings and convictions.

Can AI decide on adverse media hits?

AI can find, translate, classify and rank stories, but under Article 76 of the EU AMLR decisions to accept, refuse or end a relationship, or to change the level of due diligence, need meaningful human intervention, and customers can ask for an explanation and challenge the decision.

How often is adverse media screening done?

At onboarding and then on an ongoing basis. Many firms re-screen customers automatically and alert on new stories, with deeper manual reviews for higher-risk customers during periodic reviews.

Keep reading

• Customer due diligence vs enhanced due diligence
• Politically exposed person (PEP): meaning and rules
• Crypto watchlist and PEP screening
• AML red flags in crypto
• AML risk assessment for crypto businesses

This article is educational and is not legal advice. Legal references are to Regulation (EU) 2024/1624 (the AMLR, applicable from 10 July 2027) and 31 CFR 1020.210 as published, checked on 9 October 2026. Vendor features are described from the vendor's website on the same date. Rules change; check the current versions before relying on them.