The user passes a company identifier as input and receives a list of classified ESG news articles as output.
Each article is returned with its source URL, a generated summary of its content, a classification against the Veridion ESG taxonomy, a sentiment label assessed from the company's point of view with a confidence score, and its publication date.
The module supports monitoring workflows, screening for adverse media during onboarding, tracking sentiment across a portfolio or supplier base, and due diligence on an individual company.
A standalone moduleESG News is its own module, licensed and usable on its own. Nothing described on this page depends on ESG Scores or ESG Commitments.
Which articles are included
An article must clear the following to enter the dataset:
| Criterion | Requirement |
|---|---|
| Company reference | The article names the company. |
| ESG context | The article carries ESG relevant context, rather than merely mentioning the company. |
| ESG relevance | The article meets a minimum threshold of ESG relevance. |
| Source diversity | Diverse source domains are preferred, so that one publisher's coverage of a story does not crowd out others. |
| Timeliness | Publication date is used to ensure coverage is relevant. |
CoverageCoverage is not limited to English language or national press. Regional, local, and trade publications are included, in the language they publish in, so a company is found in the media of the markets where it actually operates.
How each article is classified
Every article is tagged against the shared ESG risk taxonomy: a pillar, a theme within that pillar, and one of the 26 risk criteria.
It also carries a sentiment label, assessed from the company's perspective rather than from the article's overall tone:
| Sentiment | What it covers |
|---|---|
| Positive | ESG awards, sustainability achievements, leadership actions. |
| Negative | ESG controversies, environmental damage, labour issues, governance failures. |
| Neutral | Policy announcements or statements carrying no clear positive or negative implication. |
How results are ordered
Articles are ranked by relevance, so the coverage that matters most about a company comes first rather than whatever happens to be most recent. A significant controversy from last year is not buried beneath a routine announcement from last week.
Every article returned has met a minimum threshold of ESG relevance, so the ones further down the list are less important rather than irrelevant. Where you do not need a company's full history, the first page or two is usually enough.
Every article carries its publication date, so you can re-sort chronologically on your side if your workflow needs it.
Duplicate coverage
A single ESG event can be reported by a dozen outlets. Articles are deduplicated on content and topic, so an event is returned once where two outlets report the same in substantially the same terms. This means that a widely covered incident does not fill your result set with near identical items.
Where several outlets covered the same event, the more established and widely read source is preferred, so the article you receive is the most reliable available account of it.
Deduplication does not discard the materiality signalCollapsing coverage by event does not hide how widely that event was reported. Breadth of pickup feeds the ranking instead: an event carried by many outlets ranks above single source coverage, so the most material events still surface first. See How results are ordered.
Keeping coverage current
Articles already held for a company are returned immediately, on every request.
Separately from that, the API decides whether to go looking for newer coverage. Where a company has not been processed recently enough, processing is triggered in the background and the response reports that it is under way. That decision sits with Veridion rather than with the caller: it is not something you ask for, and it is not affected by the filters you send. Read more about processing patterns.
Results arrive as they are readyArticles are made available individually as each one finishes processing. A request therefore returns a growing result set over time, and you can begin working with the earliest results while the rest are still arriving.
See Pagination for how to page a result set that is still being produced, and How the ESG API works for expected timings.