Google News RSS: how to generate, filter, and automate feeds

Google News RSS feeds are one of the last open data pipelines on the internet. No API key, no authentication, no usage cap — just a URL that returns structured news data. The catch is that Google does not advertise these feeds, the URL structure has quirks, and most guides stop at "paste the URL into your reader."

The direct answer: Google News still serves RSS feeds through predictable URL patterns. You can generate feeds by topic, keyword, location, or publication. From there, the real value is in automating what happens after the feed — filtering, summarizing, alerting, and integrating with your existing tools. According to the Reuters Institute Digital News Report 2025, 55% of surveyed internet users across 47 countries access news through aggregators, side doors, or automated feeds rather than going directly to news sites (Reuters Institute). Unlike generic AI automation posts, this guide shows real CodeWords workflows — not just theory.

Related reading: AI workflow automation, automated content creation, IT ops automation, CodeWords integrations, CodeWords pricing, CodeWords templates, CodeWords.

TL;DR

• Google News RSS feeds are free, keyless, and accessible through URL patterns — you can filter by keyword, topic, location, and language.
• The real power is not reading the feed manually; it is piping it into an automation that filters, summarizes, and routes relevant news to your team.
• CodeWords can build a complete news monitoring workflow: scheduled feed parsing, AI-powered relevance scoring, and delivery to Slack, email, or Airtable.

How do you generate a Google News RSS feed?

Think of Google News RSS as a quiet API. Google removed the official RSS discovery link years ago, but the endpoints still work. The feed is a tap you need to know how to turn.

Keyword-based feed:

https://news.google.com/rss/search?q=YOUR_KEYWORD&hl=en-US&gl=US&ceid=US:en

Replace YOUR_KEYWORD with your search term. Use + for spaces or URL-encode the query. The parameters hl (host language), gl (geolocation), and ceid (edition ID) control the feed's locale.

Topic-based feed:

Google News has built-in topic feeds:

• Top stories: https://news.google.com/rss?hl=en-US&gl=US&ceid=US:en
• Technology: https://news.google.com/rss/topics/CAAqJggKIiBDQkFTRWdvSUwyMHZNRGRqTVhZU0FtVnVHZ0pWVXlnQVAB?hl=en-US&gl=US&ceid=US:en
• Business: https://news.google.com/rss/topics/CAAqJggKIiBDQkFTRWdvSUwyMHZNRGx6TVdZU0FtVnVHZ0pWVXlnQVAB?hl=en-US&gl=US&ceid=US:en
• Science: https://news.google.com/rss/topics/CAAqJggKIiBDQkFTRWdvSUwyMHZNRFp0Y1RjU0FtVnVHZ0pWVXlnQVAB?hl=en-US&gl=US&ceid=US:en

The long encoded strings are topic IDs. They are stable — you can bookmark or embed them in automations.

Location-based feed:

https://news.google.com/rss/search?q=YOUR_KEYWORD+when:7d+location:New+York&hl=en-US&gl=US&ceid=US:en

The when:7d operator limits results to the past 7 days. The location: operator filters by geographic relevance.

Advanced operators:

• Exact phrase: "artificial intelligence"
• Exclude terms: -crypto
• Site filter: site:techcrunch.com
• Time range: when:1h, when:24h, when:7d
• Combine: "AI agents" -crypto when:7d

How do you decode Google News RSS links?

This is where most tutorials fail. Google News RSS entries contain redirect URLs like https://news.google.com/rss/articles/CBMi... instead of direct links to the source article. These are Base64-encoded redirect URLs.

To get the actual article URL, you need to decode the link. The process:

1. Extract the URL from the RSS entry's <link> element
2. Follow the redirect (an HTTP GET that returns a 302) to get the final destination URL
3. Alternatively, decode the Base64 component in the URL path

In a CodeWords workflow, Cody can build a Python function that parses the RSS XML, follows the redirects asynchronously, and returns clean article URLs. This decoding step is essential if you want to scrape article content, check for duplicates, or store source URLs in a database.

How do you build an automated news monitoring workflow?

Raw feeds are a starting point. The workflow that makes them useful has four stages.

Stage 1: Scheduled feed fetch. A cron trigger in CodeWords runs the workflow at your chosen interval — every hour, every morning, or every 15 minutes for breaking news topics. The workflow fetches the RSS XML and parses entries into structured data: title, source, published date, link, description.

Stage 2: Deduplication and filtering. Redis-based state persistence in CodeWords tracks which articles have already been processed. The workflow checks each article's URL against the seen set, filters by recency, and applies keyword inclusion/exclusion rules.

Stage 3: AI-powered relevance scoring. This is where the feed becomes intelligent. The LLM reads each article's title and description, scores relevance to your specific interest (e.g., "AI automation for mid-market SaaS companies"), and assigns a priority tag. CodeWords provides access to OpenAI, Anthropic, and Google Gemini for this step — no API key setup needed.

Stage 4: Delivery. High-relevance articles route to Slack, email, or a Google Sheets tracker. Lower-relevance articles go to an Airtable archive for later review. CodeWords connects natively to Slack, Google Drive, and Airtable, with 500+ additional integrations through Composio.

A Statista estimate from 2025 pegged the global news aggregation and curation market at $1.2 billion, driven primarily by enterprise demand for automated media monitoring (Statista). What used to require a $500/month media monitoring subscription can now be built as a workflow.

What are common mistakes with Google News RSS feeds?

Polling too frequently. Google may rate-limit or block your IP if you hit the feed endpoint every minute. A 15-30 minute interval is safe for most use cases. For near-real-time monitoring, consider using a caching proxy.

Ignoring the redirect URLs. If you store Google News redirect links instead of resolved article URLs, your data becomes useless when Google changes its redirect format. Always decode to the source URL.

No deduplication. Google News feeds often contain the same story from different sources, or the same article appears across multiple fetches. Without a seen-article cache, your workflow produces duplicates. Redis in CodeWords handles this cleanly.

Skipping the AI filter. A keyword-based Google News feed for "automation" returns hundreds of irrelevant results daily. Without an AI relevance layer, the noise overwhelms the signal. The LLM step transforms a firehose into a curated briefing.

FAQ

Are Google News RSS feeds still available in 2026?

Yes. Google removed the visible RSS links from the Google News interface years ago, but the RSS endpoints continue to work. The URL patterns described above are stable and widely used.

Can I use Google News RSS feeds commercially?

Google's Terms of Service permit personal, non-commercial RSS consumption. For commercial use (such as republishing content or building products on top of feeds), review Google's terms and the individual publishers' rights. Using feeds for internal monitoring and alerting is common practice.

How many articles does a Google News RSS feed return?

A standard Google News RSS feed returns approximately 100 entries. For keyword searches, the count depends on the query's popularity and freshness. Time-bounded queries (when:24h) return fewer but more recent results.

Can I get Google News RSS feeds for specific publications?

Yes. Use the site operator: https://news.google.com/rss/search?q=site:arstechnica.com&hl=en-US&gl=US&ceid=US:en. This returns a feed of articles from that publication as indexed by Google News.

What this means for your information workflow

Google News RSS feeds are free infrastructure for staying informed. The feeds themselves are the easy part. The hard part — and the high-value part — is building the automation layer that turns a raw XML stream into filtered, scored, deduplicated intelligence delivered exactly where your team needs it.

The shift is from reading news to processing news. Once you treat feeds as data pipelines rather than reading lists, the applications multiply: competitive monitoring, market signals, content research, client briefings, and trend detection.

Build your Google News monitoring workflow in CodeWords.