Kagi News: A Private, Once-Daily Global Briefing

Kagi News is a once-daily, five-minute press review that distills thousands of community-curated RSS feeds into a concise, globally diverse briefing. It prioritizes privacy (no tracking or profiling), transparency of sources, and respect for publishers by using only their chosen RSS content. Readers can customize categories and language settings, with translations handled by Kagi Translate, and access it on web, iOS, and Android.

Key Points

One daily, five-minute briefing around noon UTC delivers essential global news without endless scrolling or notifications.
Sources are community-curated, transparent, and maintained via a public GitHub repository to ensure diversity of perspectives.
Privacy by design: no tracking, profiling, or monetization of attention; users remain the customer, not the product.
Customizable experience: select and reorder categories, adjust story counts, and rearrange sections to match priorities.
Publisher-respecting implementation: uses publicly available RSS feeds only, with content exactly as publishers provide; no scraping.

Sentiment

The overall sentiment is cautiously positive but notably divided. A significant portion of the community appreciates Kagi's user-first philosophy and the once-daily publishing concept, and many paying Kagi subscribers expressed goodwill toward the company. However, there is substantial and vocal criticism around three axes: the initial lack of transparency about AI usage, concerns about LLM reliability for news, and the ethics of summarizing journalists' work without compensation. The discussion skews more favorable than a typical HN thread about an AI product, largely because Kagi has built considerable goodwill through its search product and because the founder actively engaged with criticism.

In Agreement

The once-daily publishing model is a genuinely valuable design choice that creates a natural stopping point and combats doomscrolling and news addiction
Kagi's approach to AI is the most tasteful and practical in the industry — always opt-in and focused on summarizing rather than generating content from scratch
The product fills a real gap as an alternative to algorithmic, attention-monetizing news feeds from Google News, Apple News, and social media
Kagi as a paid service aligns incentives correctly — users are customers rather than products, which builds trust compared to ad-supported alternatives
The open-source community-curated RSS feed list provides transparency and allows for continual improvement of sources
LLM summarization is a legitimate and practical use case for AI, and the summaries provide enough context to decide whether to click through to original articles
The clean, no-ads, no-images design respects users' intelligence and time, eliminating clickbait and ragebait effectively
Free availability (no subscription required for Kagi News specifically) makes it accessible as a gateway to the Kagi ecosystem

Opposed

Kagi was not transparent about using LLMs to generate the summaries, initially omitting any mention of AI in the blog post announcement
LLM-generated news summaries are inherently unreliable due to hallucination risks — citations do not guarantee accuracy and users may internalize false information
The product parasitically extracts value from journalists and news organizations without returning any economic value to the original content creators
Kagi is overextending itself across too many products (search, browser, assistant, maps, news) with only a small team, risking quality degradation
The 'all sides' framing and claim to present the full spectrum of perspectives is philosophically naive — not all perspectives are equal, and summarization inherently introduces bias
Non-LLM NLP summarization methods exist that extract key sentences without risk of hallucination and would be more appropriate for news
Once-daily updates could be a cost-cutting measure dressed up as a feature — news evolves throughout the day and important stories may be missed
Using RSS feeds as source material means the LLM may generate inaccurate summaries from misleading headlines or incomplete content
The product could create legal liability for libel if AI-generated summaries contain factual errors about real people or events
Kagi's relationship with Yandex raises concerns about neutrality and whether user funds flow to the Russian economy