This step-by-step guide is exclusively available for Lead with AI PRO membership. 🚀 With Lead with AI PRO, you’ll get: ✅ Access to expert-crafted step-by-step guides ✅ AI-powered workflows to boost productivity ✅ Exclusive tools and resources for smarter work Upgrade to Lead with AI PRO and access all premium content instantly.
ChatGPT Atlas: The AI-native browser that works with you
Plus: Let Agent in ChatGPT Atlas research and build content plan for you.
By
Evelyn Le
Strategic Product Lead, Stay Ahead, FlexOS
Presented by
Last week, OpenAI launched ChatGPT Atlas, a brand-new browser built from the ground up with ChatGPT inside. It means you can now read, search, and work online with AI always by your side.
Here’s what stands out:
Built-in ChatGPT sidebar: Ask questions, get quick summaries, or draft replies directly on any webpage, without copying or switching apps.
Works like your current browser: Import your bookmarks, passwords, and extensions from Chrome, Safari, or Edge in minutes.
Agent Mode (preview): For paid users, Atlas can handle multi-step tasks like researching competitors or comparing products on its own.
Private by design: You control what Atlas remembers. Memory is fully opt-in, and you can turn it off anytime.
ChatGPT Atlas is now available for macOS users, with Windows and mobile versions coming soon.
As Daan mentioned on Friday, Atlas hints at what the next phase of AI might look like: less about typing prompts into a chatbox, and more about AI working alongside you in familiar tools. It’s not perfect yet, but it signals a future where research, writing, and reasoning could all happen in one place.
To see what this new browser can actually do in practice, let’s put ChatGPT Atlas to work. We’ll use Atlas’s Agent Mode to handle a task that normally takes days – researching the top 10 LinkedIn creators in AI and social impact, analyzing why their content works, and turning those insights into a monthly content plan.
In just a few minutes, Agent in Atlas can search, summarize, and organize everything into a clean spreadsheet on its own.
Here’s how to make it happen, step by step.
Flagship AI Newsletter
The AI Newsletter That Makes You Smarter, Not Busier
Join over 30,000 leaders and receive our insights on AI platforms, implementations, and organizational change management.
In 5 Steps: Let ChatGPT Atlas Do the Research and Build Your Content Plan
Step 1 – Set up ChatGPT Atlas
Atlas works like your regular browser, but you’ll first need to connect it to your ChatGPT account and turn on Agent Mode. This gives it the ability to browse the web, collect information, and organize it for you.
Sign in using the same email as your ChatGPT account.
In the top-right corner, toggle Agent Mode (Preview) on. This unlocks multi-step actions like data gathering and file creation.
Then paste this prompt into the sidebar:“Browse LinkedIn to find the top 10 creators in AI and social impact over the last 30–60 days, list them with name, URL, followers, niche, and posting style, then collect their top-performing posts (topic, format, summary, why it worked) and create a Google Sheet with 2 tabs — Top Creators and Insights Hub. The Insights Hub should include top posts, patterns, gaps, and a 4-week content plan.”
When you first launch the Agent, it’ll ask a few questions about privacy settings. You’ll then need to log in to your LinkedIn and Google accounts to give it access. Once you hit Enter, your Agent starts exploring LinkedIn on its own.
Step 2 – Let the Agent do the heavy lifting
Once the prompt is in, Agent takes over. It acts like an analyst who browses, extracts, and organizes, all while you focus on higher-level thinking.
Actions:
Agent will search on the web for the top 10 creators in AI and social impact.
It will visit each creator’s profile on LinkedIn and identify their best-performing posts by engagement. You’ll have to take control and sign in to your LinkedIn account at this stage.
It will create a Google Sheet with two tabs:
Top Creators – names, niches, and style notes.
Insights Hub – a database of top post topics, summaries, why it worked, content patterns, gaps, and a content plan for you for the next 4 weeks.
You’ll see in real time how it works, screen creators’ posts, and build the sheet.
Step 3 – Refine with real-time feedback
The magic of having the ChatGPT Agent right in your browser sidebar is how collaborative it feels. You can correct, guide, or reshape its work as it happens.
Actions:
Review the Google Sheet as it’s being built.
Give direct feedback such as:
“This doesn’t seem right, please try again.”
“Let’s put these columns in a separate tab.”
Atlas will instantly edit and reformat based on your messages without rewriting prompts from scratch.
Step 4 – Enhance your content
With the foundation built, now comes the creative payoff, which is transforming analysis into action.
You can:
Ask Agent to generate draft captions for each content idea inspired by top-performing posts.
Refine the tone and voice so the writing feels more aligned with your style and expectations.
Request content pillars based on observed patterns.
Create a visual brief for each content idea.
It can even help you repurpose ideas into carousels, thought pieces, or discussion prompts.
Step 5 – Experiment and scale your workflow
Once you see how smoothly ChatGPT in Atlas manages research and planning, you can expand the workflow.
Ideas to try next:
“Research the top AI products launched this quarter and categorize them by business use case.”
“Track competitors’ messaging across their websites and social channels to identify positioning gaps.”
“Analyze trending topics across TechCrunch, The Verge, and X to inform our next thought-leadership angle.”
“Build a reading dashboard that summarizes industry news, reports, and funding updates in one place.”
“Compare my own LinkedIn profile to these top creators and identify improvement areas.”
“Turn the 4-week plan into a publishing calendar with dates and formats.”
“Summarize engagement patterns into a one-pager for my team.”
“Set up a recurring weekly task to refresh the dataset.”
You’ll end up with a self-updating system that learns from the best and keeps your content strategy sharp, relevant, and informed.
For over a year, Lead with AI Boot Camp has helped more than 1,000 global leaders design, build, and deploy their own AI assistants. This November, we’re running the last cohort in this current format.
This is your final chance to join the program loved by 1,000+ global leaders.
In 3 weeks, you’ll ship 5 functional AI Assistants tailored to your workflow and build a real, deployable AI micro-app during the Agentic Vibe Coding Workshop.
Why leaders choose this Boot Camp:
Proven impact. More than 1,000 global leaders from NVIDIA, Unilever, Microsoft, and more have completed this exact format
Always current. Includes the latest breakthroughs in AI so you always stay ahead.
Hands-on, not passive. Live sessions, small-group chats, and 1:1 mentor support ensure you ship, not just learn.
Real outcomes. Graduate with working AI assistants and a recognized certificate in applied AI leadership.
Last opportunity to experience Lead with AI Boot Camp. Seats are 50% gone. Enroll today.
Want to reach 30,000+ business leaders applying AI in their work, teams, and organizations? Advertise with us.
Your AI Team: Claude Skills and Microsoft 365 connectors, Google’s Veo 3.1, Gemini’s meeting scheduling in Gmail, and more.
Every week, I report on the top updates to your favorite AI tools.
This week, it's definitely the era of agents:
Google’s Veo 3.1 updates in Flow
Google announced an update to its generative video ecosystem: the release of the Veo 3.1 model and new creative tools inside its filmmaking platform Flow. Here are the key takeaways for senior executives and strategic decision-makers:
Richer audio + enhanced realism: Veo 3.1 brings audio generation that’s synchronized with video, such as dialogue, ambient sound, SFX, and tighter fidelity in visuals and narrative control.
More granular creative control: Inside Flow, you get features like:
“Ingredients to Video” (plug multiple reference images to direct characters, objects, and style)
“Frames to Video” (specify a start and end image and let the model generate the transition)
“Extend” or “Scene Extension” (take the last second of a clip and generate additional time/seamless continuation)
Insert & Remove Objects: Flow allows for editing of generated videos by adding new elements (“Insert”) or removing undesired ones (“Remove”) with reconstructed background/lighting handling.
Veo 3.1 is accessible both via the consumer app and through the Gemini API and Vertex AI for developers/enterprises.
Claude now integrates with Microsoft 365 (Outlook, Teams, OneDrive, SharePoint) through Anthropic’s MCP (Model Context Protocol) connector, enabling it to search and analyze documents, email threads, and chat conversations directly from within your existing productivity stack.
This opens a strategic opportunity for your organization, enabling Claude to surface insights across large internal datasets and become an operative part of your knowledge workflows.
Anthropic has released its latest lightweight model, Claude Haiku 4.5, which delivers performance comparable to its larger Sonnet 4 model at roughly one-third the cost and more than twice the speed.
Designed for high-throughput scenarios, Haiku 4.5 can be spun up in parallel, making it ideal for large-scale deployments in workflows like code generation, document synthesis, and long-context tasks.
Anthropic introduces “Skills” for Claude to tailor enterprise workflows
Anthropic is launching a new modular feature called “Skills” that allows developers and business teams to package instructions, scripts, and resources into self-contained bundles that the Claude agent will load when appropriate.
This makes Claude far more applicable in enterprise settings, whether automating spreadsheets, enforcing brand guidelines, or embedding domain-specific logic, without needing to retrain the core model.
Google Gemini in Gmail: “Help me schedule” makes meeting scheduling easy
Gmail has added a new Gemini-powered feature, “Help me schedule,” which automatically detects when you’re setting up a meeting in an email, suggests available time slots based on your Google Calendar and the email’s context, and inserts them into the message for the recipient to pick from.
Once the recipient chooses a slot, the invite is created for both parties with no back-and-forth.
The AI Newsletter That Makes You Smarter, Not Busier
Join over 30,000 leaders and receive our insights on AI platforms, implementations, and organizational change management.
Exclusive for PRO Members: What's The Fuss About Claude Code?
Hosted by Helen Lee Kupp, founder of Women Defining AI, this hands-on workshop is for leaders who are curious about Claude Code and want to understand how it can be used to think and build systems in powerful new ways.
Helen will walk you through:
Introduction to Claude Code: What it is, when to use
Demo real use cases, including non-technical ones
How to get started with the basic technical setup - an overview of using the terminal, GitHub, and setting up Claude Code, and what’s involved
Live Q&A to help you get unstuck and build confidence
If you’ve wanted to try your hand at more technical tools or simply want a safe space to ask questions, this is the perfect place to start.
The session takes place on November 13th at 7PM London / 11AM Pacific Time/ 3PM Eastern Time.
Members, the invite is on your calendar.
Not a member yet and want to join this exclusive Masterclass?
Or get a 14-day free trial of the world's leading executive AI community here.