Editorial Policy

What's the Big Data publishes independent reporting, analysis, and how-to guides covering artificial intelligence, big data, machine learning, and the companies building them. This page explains how our newsroom operates: how we pick stories, how we verify facts, how we handle AI-assisted writing, and how we treat corrections.

Editorial independence

Our editorial team decides what to cover and how to cover it. No advertiser, sponsor, tool vendor, or affiliate partner has any say in the assignment, angle, tone, or timing of a story. Sponsored content, when it appears, is clearly labelled as such and is written or reviewed by our sales and marketing team — never by reporters covering the same beat.

How we choose what to cover

We prioritise stories that are timely, verifiable, and useful to readers who work with or make decisions about AI. On the news desk (/ai-news/) that generally means announcements from major model developers, regulator actions, funding and M&A that will change the market structure, first-hand incident reporting, and research whose conclusions the wider industry is likely to act on.

Sourcing standards

Every news article on this site is expected to link to at least one primary source (an official company blog post or press release, a filed regulatory document, a peer-reviewed paper, a direct interview, or a public statement from a named individual). When we rely on reporting from another outlet, we say so in-line and link to their original coverage — we do not launder second-hand claims as our own. Rumours, unverified leaks, and anonymously sourced screenshots are labelled as such and are only published when the underlying question is materially in the public interest.

Fact-checking and review

Before a news story is published, the writer verifies each factual claim against the primary source and an editor reviews the piece for accuracy, framing, and clarity. Technical claims (benchmark numbers, model capabilities, revenue figures) are checked against the original filing or announcement rather than a summary. If we can't verify a claim, we either omit it or state clearly that it is unverified.

Use of AI in the newsroom

We may use AI tools to summarise long documents, transcribe interviews, brainstorm headline options, or draft short factual captions. Every published article is written and edited by human journalists, who are responsible for accuracy and framing. We do not publish articles that are wholly AI-generated without human editorial oversight, and we do not use AI to fabricate quotes, invent sources, or generate photorealistic images of real people. When an image is AI-generated, the caption says so.

Bylines and author identity

Articles carry the byline of the human journalist who wrote them. Author pages (linked from each byline) show every article that author has published on the site, along with a short professional biography where available. If more than one journalist substantively contributed to a story, we credit the primary author in the byline and acknowledge additional contributors at the end of the article.

Timeliness and updates

News moves fast, especially in AI. When a story we've published develops materially — new numbers, a corrected figure, a company response — we update the article in place, add an “Updated” timestamp near the byline, and (for corrections) note what changed at the bottom of the piece. See our Corrections Policy for the specifics.

Diversity of voices

We actively look for sources, analysts, and researchers whose perspectives are under-represented in mainstream tech coverage, including women, researchers based outside the United States, and practitioners working in applied domains outside the largest model labs. This is a work in progress and we invite readers to flag stories where our sourcing has fallen short.

Reader feedback

If you spot an error, disagree with our framing, or think we missed important context, please email [email protected] with the article URL and the specific point you want us to address. We read everything and reply to correction requests first.

Related policies

Last updated: September 2026.