What a Personal AI News Agent Actually Is
The term "AI agent" is everywhere right now — and, as covered in the previous post on this site, it means something specific. An agent is not a chatbot you ask questions. It is an automated workflow that takes action on your behalf, on a schedule, without you doing anything manually.
A personal news agent is one of the simplest and most immediately useful examples of this. You set it up once. Every morning — or every week, if you prefer — it wakes up, asks an AI model to research and summarise the topics you care about, and sends the result to your inbox. No prompting, no copying, no searching.
The distinction matters because most professionals use AI reactively — they open a tool, type a question, and get an answer. An agent flips that. It does the work before you even know you need it.
This is not a complicated build. If you can fill in a form online, you can build this. The entire setup takes under 30 minutes and requires no technical background whatsoever.
Which AI Model to Use
The automation platform used in this build — Make.com — connects natively to all four major AI models. The right choice depends largely on how you already work.
| Model | Best for | Why it works here |
|---|---|---|
| Claude | Clear, well-structured writing | Produces clean, readable briefings with excellent formatting |
| Perplexity | Real-time news and research | Search-native — pulls live information by default |
| Gemini | Google Workspace users | Natural fit if your workflow already lives in Google |
| ChatGPT | Broad all-purpose output | Most familiar to users already on OpenAI tools |
For a news-focused agent, Perplexity has a structural advantage — it is built around live search and cites sources by default. Claude and ChatGPT produce more natural prose. The build steps below work identically regardless of which model you choose. Only the module name inside Make.com changes.
How the Automation Works
The agent is a three-step chain. A scheduler fires at a time you choose. It sends your prompt to an AI model. The model's response is delivered to your inbox. That is the entire mechanism.
Make.com is the automation layer that connects everything. It acts as the bridge between the scheduler, the AI model, and your email. You do not need to write code, manage APIs, or touch anything technical. The entire setup is done through a visual interface where you connect modules like building blocks.
The Prompt That Powers It
The quality of your briefing depends almost entirely on the quality of your prompt. A vague prompt produces a vague briefing. A well-structured prompt produces something genuinely useful every morning.
Here is the prompt used to build and test the agent described in this post. Copy it exactly, then adjust the city and sections to suit your own context.
☀️ London Weather
Today's weather in London — temperature, conditions, and anything worth knowing before leaving the house.
📰 UK & World News
The 3 most important news stories from the UK and around the world today. For each: one sentence on what happened and why it matters.
💻 Tech News
The 3 most significant technology stories today. Keep it jargon-free — explain why each story matters to a professional, not a developer.
🚀 Productivity Tools to Watch
1 to 2 new or newly updated productivity apps or AI tools worth knowing about this week. For each: what it does, who it's for, and whether it's free.
Keep the entire briefing under 400 words. Plain text only — no markdown, no bullet symbols, just clean numbered lists and clear headings. Write as if briefing a senior executive who has 2 minutes to read this.
A few things to note about how this prompt is written. It tells the model its role upfront, specifies the format precisely, sets a hard word limit, and defines the audience. Each of these constraints removes a decision the model would otherwise make on its own — and that is where briefings become inconsistent over time.
The prompt is fully editable. If you work in financial services, replace the tech news section with a markets summary. If you cover a different geography, change the location. The structure is the template; the content is yours to define.
Step-by-Step Build Guide
This build uses Make.com's free tier and takes approximately 25 minutes. You will need a Make.com account and a Gmail account. No other tools are required to get started.
Go to make.com and sign up. The free plan allows two active scenarios and 1,000 operations per month — more than enough for a daily briefing running 30 times a month.
From your dashboard, click "+ Create scenario." This opens the visual scenario builder — a blank canvas where you will add your modules.
Click the large "+" button on the canvas. Search for your chosen model — "Anthropic Claude," "Google Gemini AI," "Perplexity AI," or "OpenAI" for ChatGPT. Select the "Simple Text Prompt" action (or equivalent). Choose the most recent available model version from the dropdown, then paste the prompt from the section above into the Text Prompt field. Save the module.
Click the "+" that appears to the right of your AI module. Search for "Gmail" and select "Send an email." Connect your Gmail account when prompted. Set the To field to your email address, the Subject to "Daily Brief: Weather, News, Tech," and in the Body field, map the output from your AI module by clicking inside the field and selecting the Result variable from your AI module. Save.
At the bottom of the screen, click "Every 15 minutes" and change it to "Every day" at your preferred time. 6:30am works well — the briefing lands before most working days begin. You can change this to weekly or monthly at any point.
Click "Run once" to trigger the scenario manually. Check your inbox — the briefing should arrive within a minute. If it does, click "Activate scenario" to switch the daily schedule on permanently.
Working across large organisations at Microsoft, Expedia, and Citi, I watched teams spend time each morning piecing together context before they could start actual work. This build replaces that entirely — and runs without anyone touching it.
Make.com — the automation layer that powers this build
Make.com connects your AI model to your email, calendar, Slack, and 1,500+ other apps — no code required. The free plan is generous enough to run a daily briefing indefinitely.
Start free on Make.com →What to Expect From the Output
The first few runs will give you a clear sense of what works and what needs adjusting. A few things worth knowing before you start.
Real-time accuracy varies by model
Models with built-in web search — Perplexity in particular — consistently pull same-day news. Models without live search access produce more general summaries, particularly for breaking news. If current information matters to you, choose Perplexity or ensure your chosen model has web browsing enabled in its settings.
The prompt is your control lever
If the briefing is too long, tighten the word limit. If a section is not useful to you, remove it. If you want a specific industry focus — financial services, healthcare, or retail, for example — add that context to the prompt. Every adjustment takes 30 seconds and takes effect on the next run.
You can add more outputs later
Once the email version is working, it is straightforward to add a second output — a Slack message to a personal channel, or a WhatsApp message via the WhatsApp Business API. The AI module and the prompt stay the same; you simply add another module to the end of the chain.
For more context on how AI agents fit into a wider professional workflow, the introduction to AI agents on this site is a useful starting point.
Frequently Asked Questions
A personal AI news agent is an automated workflow that runs on a schedule, uses an AI model to research and summarise current news, and delivers the output to you — by email, WhatsApp, or another channel — without any manual input from you.
No. The build in this post uses Make.com, a visual no-code automation platform, and an AI model of your choice. No coding is required at any point. If you can fill in a form, you can build this.
All four major models — Claude, Gemini, Perplexity, and ChatGPT — connect to Make.com. Perplexity is particularly strong for news-focused agents because it is search-native and pulls live information by default. Claude and ChatGPT produce well-structured, readable output. Gemini is a natural choice if you already use Google Workspace.
Yes — the prompt is fully editable at any time inside Make.com. You can change the sections, adjust the tone, add a specific industry focus, or change the format entirely. You can also change the schedule from daily to weekly or monthly in a single click.
The Practical Takeaway
This agent takes under 30 minutes to build and will save you that time back every single morning. The value is not just the briefing itself — it is the habit it replaces. Instead of opening four browser tabs, scrolling a news app, and piecing together a picture of the day, you get two minutes of structured, relevant information delivered before you start work.
More importantly, this is the simplest possible introduction to what AI agents actually do. Once you have built this, the logic applies everywhere — monitoring client emails, tracking competitor activity, summarising weekly reports. The mechanism is identical. Only the prompt and the output change.
One AI tip, every Tuesday.
Practical AI agent ideas for professionals — no jargon, no hype. Join the Linda & AI newsletter.