Digital Influence is written by AI. We say so on every post, and this page explains how the work is done, what is checked before anything publishes, and where people are involved.
Editorial standards and AI disclosure
How a story is found
Our research desk watches around 90 sources every 15 minutes: company newsrooms, trade publications, national press, regulators and research groups, in English, Spanish and Portuguese. Each new item is scored by a classification model for how directly it affects commerce teams and how significant it is. Items that report the same development are grouped, so one event becomes one post. Most items are not covered; the desk keeps a log of every decision and the reason for it.
How a post is written
For each item that clears the bar, the desk reads the original article and an AI model (Claude, made by Anthropic) writes the post: a summary of what was reported, Our take, and key takeaways. The model works only from the source article and other coverage of the same development. It is instructed never to add facts, numbers or quotes that are not in the source.
When the original is behind a hard paywall, the desk waits for an openly available report of the same development and writes from that, crediting it as the source.
What is checked before a post publishes
- Attribution. If a source attributes a claim or hedges it ("according to", "expects", "unverified"), the post keeps that attribution and hedge. A company's claim is never turned into a plain statement of fact.
- Fact check. A separate model checks every statement in the headline, summary, takeaways and Our take against the source article. If a statement cannot be confirmed, the post is rewritten once; if it still cannot be confirmed, it does not publish and goes to an editor.
- House style. Every post must pass our copy rules: plain, specific language, no hype words, American English.
- Safety. Article text is screened for hidden instructions aimed at AI systems before the writing model reads it.
Posts that pass every check publish automatically. Our editors spot-check published posts and correct or remove them when needed.
Analysis and DI Daily
Analysis pieces are researched with web search, drafted by AI, fact-checked against their sources, and approved by an editor before they publish. Each one lists its sources.
DI Daily is written each weekday morning from the posts published since the previous edition. It links every development to its post and to the original reporting, and goes through the same copy and fact checks.
Our take
Our take explains why a development matters to commerce teams and what to watch, ask or do. It aims to help you form an informed position, not to persuade. It is not legal, financial or investment advice.
Sources and images
Every post names its source and links to the original reporting; additional coverage is listed below each post. We summarize; we do not republish articles. Lead images come from the original publisher and are credited on the post. Publishers who would like an image removed can tell us through our contact form and we will take it down.
Independence
Coverage decisions are made by the research desk under the rules on this page. No one pays to be covered. When the site carries advertising, ads will be clearly separate from editorial content.
Corrections
If you find an error, report it through our contact form and include the post's address. We correct mistakes promptly and note significant corrections on the post.