About
A wire service for the frontier.
The Frontier Wire reports on artificial intelligence for people who build with it, buy it, regulate it or simply need to keep up. We publish what changed, why it matters, and what the evidence actually supports.
What we cover
The publication is organised into six desks. Each has its own page and RSS feed.
- News: What happened in AI this week: model releases, funding, policy, and the people behind them.
- Research: Briefings on the papers, benchmarks and techniques that move the field, explained without the jargon.
- Tools: Reviews and field notes on the models, frameworks and products people actually build with.
- Comparisons: Side-by-side evaluations of competing tools and models, with the criteria spelled out.
- Editorials: Opinion from our editors and guest writers on where AI is going and where it should not.
- Surveys: Landscape surveys and state-of-the-field reports that map an entire category at once.
Editorial principles
Evidence first. Claims about model capability are tied to a benchmark, a paper or a reproducible test we ran ourselves. When a number comes from a vendor, we say so.
No sponsored rankings. Comparisons and surveys are never paid for by the companies in them. Criteria are published alongside the verdict so you can disagree with the weighting.
Opinion is labelled. Editorials live in their own section and are marked as opinion in the page's structured data, so search engines and readers can tell them apart from reporting.
AI-assisted, human-edited. We use AI tools in research and drafting. Every published article is edited, fact-checked and signed off by a named editor.
Corrections
If we got something wrong, write to editors@thefrontierwire.com. Material corrections are noted at the top of the article with the date; the article's updated date changes as well.
Contact and tips
Story tips, embargoed briefings and reader mail all go to editors@thefrontierwire.com. Machine readers can start from llms.txt.