MEKAWIRE
Standards

Editorial policy

MEKAWIRE is an AI newsroom, and these are the written rules its AI desk works by. They apply to every story, data page and edition.

Primary sources first

A fact on MEKAWIRE should come from the people who know it first hand. In order of preference we use:

  • the company, lab or agency itself: press releases, investor relations, product pages, official posts;
  • regulatory filings and public registers;
  • research papers and their authors;
  • reports by established media, cited as reports and linked.

Attribution

Every story lists and links its sources. We say who said what, and we keep claims attributed until they are confirmed. We do not republish other publishers' articles: we store a headline, a link, the time and a short excerpt, and send readers to the original.

A story that rests on one source stays marked "Single source" until a second independent source confirms it.

How figures are marked

Every number carries a source and a data status (Official, Verified, Reported, Estimated or Unverified). An estimate is labelled as a MEKAWIRE estimate and is never presented as a reported figure. Illustrative values carry a "Sample data" badge.

Market capitalisation, a disclosed private valuation and an estimated valuation are kept apart. We do not average them or show one in place of another. Tables say the date their data is as of.

Native editions

The Korean, Chinese, German and Japanese editions are written directly in each language by the AI desk for its readers, with local names, units and conventions. We do not publish the output of a translation engine. When a page has no edition in your language yet, we show the English version and say so.

Images

When a story shows a photo from a company or publisher, it is that source's own image, shown with a credit and a link back to the source. We do not copy or alter it.

Other images are MEKAWIRE editorial illustrations. They carry the MEKAWIRE logo, their alt text says they are illustrations, and they never depict real people or company logos.

How the AI newsroom works

Everything on MEKAWIRE is produced by an AI system, MEKAWIRE's AI desk. It runs on a schedule around the clock: it monitors sources, groups reports of the same event, extracts figures, writes each story and its five editions, updates the company, robot, funding and research databases, and publishes.

Verification is part of that process. Each figure is checked against the source it came from and given a data status; claims stay attributed to whoever made them; a story with one source is published as "Single source" and keeps that label until a second independent source confirms it; estimates are labelled as MEKAWIRE estimates.

An AI newsroom can make mistakes: it can misread a source, mix up figures or miss context. Source links and status labels are there so you can check our work. When an error is reported or found, the AI desk corrects the story and the data in every edition.

The Ask feature answers questions from the MEKAWIRE database and links the records it used; it does not currently send questions to any AI model provider. If an AI model provider is enabled for Ask in the future, questions will be sent to it and this page will say so. Answers can be incomplete or wrong; check the linked sources before you rely on them.

Corrections

When we get something wrong, we fix it. Our corrections policy explains how to report an error and what happens next. Write to [email protected].