Is news about AI getting more positive or negative?
The first edition is being prepared. It will appear here with its figures, its trend and the data to download.
Why we track AI news tone
Is the news about AI getting more hopeful or more worried? This report measures the tone of English-language news about artificial intelligence every month, using the GDELT Project, which scores news articles from around the world, and counts how much of the news AI takes up.
It shows the mood a business owner is reading every day, which shapes what customers and staff expect. It is not public opinion - for that, see the UK government public attitudes to data and AI tracker. Read it beside our UK Business AI Adoption report and our guide to keeping up without chasing launches.
Every edition of AI news tone
Each edition is frozen when it is published; a later fix is a dated correction on that edition, never a silent change.
None published yet.
How AI news tone is measured
Every English-language news article the GDELT Project monitors that mentions "artificial intelligence" is scored for tone: the share of positive words minus the share of negative words, so 0 is neutral, above 0 leans positive and below 0 negative (most news sits between -5 and +5). We take GDELT's daily tone and daily article counts, and weight each day by its articles to get a month's tone. Share of coverage is those articles as a share of every article GDELT monitored. Hacker News stories are the stories posted there that mention "artificial intelligence", counted through its public search.
What it does not tell you
- Tone is measured by counting words, not by reading meaning: a story about an AI risk written calmly can score neutral.
- It covers English-language news only, and only the phrase "artificial intelligence" - stories that say only "AI" or name a product are not counted.
- It describes what the news says, not what the public thinks; for that, see surveys such as the government's public attitudes tracker.
Method versions
- Version 1, 2026-10-26
Sources
How to cite AI news tone
Quote a figure with the edition it came from - each edition has its own cite line. To cite the report as a whole, use this.
Eclipse Labs AI (2026). AI News Tone. https://eclipse-labs.ai/reports/ai-news-tone. Licensed under CC BY 4.0.
Tone and article counts: The GDELT Project (gdeltproject.org). Story counts: Hacker News search by Algolia.
Questions about AI news tone
It is the share of positive words minus the share of negative words in news articles that mention artificial intelligence, averaged over the month. Zero is neutral, above zero leans positive and below zero leans negative. Most news sits between -5 and +5, so a move of half a point is large. The GDELT Project scores every article the same way.
Read the newest edition on this page: it gives the month's tone, how it moved since the month before and a year before, and the trend since we began tracking. The score describes the words news uses, not whether AI is good or bad. For what AI can and cannot do in practice, read what AI can't do for your business.
No. It measures the tone of news articles, which is what journalists write, not what the public thinks. Public opinion needs a survey of people; the UK government runs one, its public attitudes to data and AI tracker. We keep the names apart on purpose, so a figure from this report is never quoted as public opinion.
Hacker News is where many software engineers share and argue about technology, so the number of AI stories posted there each month is a rough signal of how much the people building with AI are talking about it. We publish the count only, never the posts or who wrote them. Our reports explain every source.
Yes. Quote the figure with the edition it came from and a link; each edition has a cite-this line and a CSV, free to reuse under CC BY 4.0. GDELT asks that its data is credited with a link, and the attribution line on each edition does that. See every report for the others we publish.