AI Use Declaration
The First Click calls itself a publication for the age of AI. That should mean something concrete. This declaration states exactly where artificial intelligence is used in producing what you read, where it is not, and the rules that govern it. We would rather over-disclose than let you wonder.
1. The principle
AI assists. Humans decide. Every editorial decision on The First Click — what to cover, what to say about it, and whether to publish — is made by a human being. AI systems in our pipeline are tools our editors direct, in the same way earlier newsrooms used wire feeds, search, and databases. No content reaches you without human review and approval.
2. Where we use AI
We use AI systems, under human supervision, to:
- Monitor and gather — scanning a large number of news feeds and sources continuously, far more than a human team could watch alone;
- Classify and organise — sorting incoming stories by topic, region, and relevance so editors can find what matters;
- Extract and connect — identifying the people, organisations, and places in a story and linking them to prior coverage, which powers the context and history we attach to the news;
- Summarise as a starting point — producing first-draft summaries of source material that editors verify against the original, edit, or discard;
- Rank and surface — helping order the day's stories by significance, using criteria set by our editorial team;
- Check our work — flagging possible inconsistencies, missing attributions, or claims that need verification.
3. Where we do not use AI
- No autonomous publishing. Nothing is published, emailed, or pushed to readers by an AI system acting alone.
- No AI editorial judgment. AI does not decide what deserves coverage, what a story means, or what our analysis concludes.
- The editor’s note is written by a person. The short note beneath a story’s summary — why we chose it and what it means for you — is composed by the editor who made that choice, on the day, and is never drafted, suggested or generated by a model. The summary above it may begin as a machine draft, as section 2 says; the note never does.
- No synthetic sourcing. We do not use AI to invent quotes, sources, data, or events. A quotation on The First Click is something a person actually said, from an identified source.
- No fake imagery. We do not publish AI-generated images presented as photographs of real events or real people. Any illustrative AI-generated visual, if used, is labelled as such.
- No undisclosed AI authorship. If a piece of writing is substantially machine-generated rather than human-written, we label it. Our standard practice is that published prose is written or fully reworked and verified by our editors.
4. Accuracy and the limits of these tools
AI systems make mistakes — they can mis-summarise, miss context, or state falsehoods fluently. Our safeguards are: verification of AI-assisted drafts against original sources before publication, human approval on everything, structured validation of machine outputs in our pipeline, and a public corrections process. If an error traceable to an AI tool gets through, we correct it the same way we correct any error — visibly — and we fix the process that let it through.
5. Your data and AI
We do not use readers' personal information to train AI models, and we do not permit our technology vendors to train their models on our readers' data. Where AI helps personalise what you see, it works only from your reading activity on our platform, under rules set by our editorial team, as described in the Privacy Policy.
6. Respect for original journalism
Our AI-assisted monitoring reads publicly available reporting in order to point readers to it, with attribution and links — the same courtesy we ask of others. We do not use our systems to launder other outlets' journalism into unattributed content. Conversely, our Terms of Service prohibit the use of our content to train machine-learning models without our written permission.
7. Disclosure in practice
Where AI played a role beyond the standard pipeline described above — for example, an AI-assisted data analysis underpinning a piece — we say so in the piece itself. Our newsletter and site content is produced by the pipeline described in section 2 and reviewed by human editors; this declaration serves as the standing disclosure for that process.
8. A living document
The tools will change; this page will keep up. We review this declaration quarterly and whenever our use of AI materially changes, and we update the date at the top when we do. Meaningful changes are announced to members.
Questions about our use of AI: editor@the-first-click.com.
