Global and local languages, together
A search allows a name to be screened in a maximum of three languages, with both global and local sources being checked at the same time so that a story which had appeared only in the local press is not missed.
Adverse media screening runs in over 190 countries, using both global and local languages. The searches run in real time across news and social media sources, and then the models filter the results to flag stories that pose a real risk. The findings are classified by risk type.
The adverse media screening fails in three ways. First, it looks for the story in English even when it was published in a local language. Second, it flags every negative article, so a bad product review ends up next to a fraud indictment. Third, it returns results without the source an auditor would want to see. Our company runs real-time searches and applies keyword rules before using models to retrieve and sort results into sector-appropriate categories.
A search allows a name to be screened in a maximum of three languages, with both global and local sources being checked at the same time so that a story which had appeared only in the local press is not missed.
Include or exclude social media, limit searches to reliable sources, or decide what your team considers a trusted source. Make sure that the irrelevant information is kept out and the actual risk is clearly visible.
Each hit contains the source article, the link to that article, and the passage that is flagged. Reports that are to be reviewed or audited are produced from the same record.
complead scans news, social media, and sources in local languages worldwide, processing more than 300,000 articles each day. During each screening, it also checks sources in real time, so the search includes the most recent stories rather than only archived ones.
Stories are matched to the person or company you are screening based on the relevance rules you set, then a model reviews each match to eliminate namesakes or mere references and ensure only genuine positives reach your team.
The crime or misconduct described in each case is used to assign it to a risk category, for example, fraud, bribery, financial crime, or organized crime, rather than the article's tone. Categories can be tailored to your sector.
Each hit comes with its source, category, and evidence, delivered through the complead platform or via the API, so your team can review it, add it to a case, or include it in a report.
Over thirty languages are read, with as many as three being handled in one screening. The stories are located in the language in which they were originally published, not just in English.
The cases are arranged by the type of crime or misconduct they involve, for example, fraud, bribery, financial crime, or organized crime, not by sentiment. Category sets can be tailored to match your sector.
Each hit includes the source article, the link, the category, and the passage that triggered it; findings remain valid during an audit and can be exported as a report.
The models ranked highest eliminate duplicates and references that are only mentioned briefly. Your reviewers therefore see fewer results, and these are of a higher quality.
You can get news, social media, and local-language sources in one feed. Turn the social media feature on or off, or adjust the settings to include only the sources you trust.
When new stories are published, the system rechecks the names you're monitoring and sends your team a new hit with its category and source, no manual search needed.
The story linked with the entity, including its risk category, source, and language.
The passage that caused the hit has been retained for the purpose of review and audit.
The rules, categories, and sources you select mean only the most important stories reach the reviewer.
News, social media and local-language sources, read in 30+ languages across 190+ countries, with every hit placed in a risk category and kept with its source.
When adverse media appears on the same Fusion customer record as sanctions and PEP status, along with the customer's risk category, the risk score reflects the subject of the story, not the number of stories.
Provide a name or a company name, and you will receive relevant results, each with its risk category and source. You can open the evidence for any result with a single request.
News, social media and local-language sources across 190+ countries. You can narrow the feed to the sources your team trusts.
More than 30 languages are read, and one screening can search a name in up to three of them at once. Global and local sources are checked together, so a story that ran only in the local press still surfaces.
By the crime or misconduct a story describes, such as fraud, bribery, financial crime or organized crime, and not by its tone. A bad product review and a fraud indictment never land in the same place.
Yes. Category sets can be adjusted to match the risks your sector and your policy care about with the customizable keywords.
Stories are first matched against the relevance rules you set. A model then reviews each match and removes namesakes and passing mentions, so only genuine positives reach your team.
The matched entity, the risk category, the source and date, the language of publication, and the passage that triggered the hit, with a link to the original article.
Yes. The original-language passage stays alongside the translation, so any decision can be checked against what was actually published.
More than 300,000 articles are crawled every day, and every screening also runs real-time queries. Results include the latest stories, not just the archive.
On Fusion, adverse media sits on the same customer record as sanctions and PEP status. The score reflects what a story is about, not how many stories there are.
Yes. Each hit keeps its source, link and flagged passage, and review and audit reports are produced from that same record.
We went from checking customers one by one to a single platform that screens more than 3,000 of them around the clock, so our analysts can finally focus on real risk.
We moved from screening customers one by one to a unified platform where our analysts can focus on what actually matters.
We focus on real risks, not false positives, meeting our AML obligations and our customers' expectations.
Provide us with a list of names. During a 30-minute session, we will screen these names in the languages that are required and display the appropriate stories, sorting them by category and indicating their sources.