Rafomac News, Tech & Trends That MatterOpenAI has introduced a game-changing yet controversial new feature: ChatGPT can now track everything you do on your Mac desktop. This computer history feature captures your work activity across apps and websites, enabling AI to summarize and automate tasks without you repeating yourself.
ChatGPT Watches Over Your Shoulder
OpenAI has rolled out a new feature in its ChatGPT desktop app for Mac that takes AI assistance to the next level. Called computer history, it records user activity across apps and websites, enabling ChatGPT to recall and summarise what you’ve been working on—without you having to explain or repeat your tasks. This isn’t just memory; it’s an AI that literally watches your workflow unfold on your screen.
This feature is strictly limited for now: only Mac desktop users on a pro, business, or enterprise plan can access it, which means a hefty price tag of at least $100 a month. Moreover, it’s unavailable in Europe, the UK, and Switzerland due to privacy regulations, so expect a careful, phased rollout.
Powerful Automation Meets Privacy Risks
Imagine finishing a complex video upload workflow and instead of painstakingly explaining every step to ChatGPT, you simply ask it to create a reusable skill from your recent activities. With computer history on, ChatGPT automatically captures those steps in real-time, turning them into workflows you can repeat effortlessly.
But this power comes with an unsettling price. ChatGPT is essentially spying on everything you do on your machine—from texts and emails to Zoom calls and browsers—unless you explicitly exclude certain apps. OpenAI advises users to exclude sensitive applications like finance or health apps and to pause tracking during communications to respect privacy and legal boundaries. Yet, trusting an AI to chronicle your entire desktop activity is a big leap.
How It Works—and What It Sees
The setup lets you handpick which apps ChatGPT tracks. One user excluded messaging and some internal tools, whittling the list down to essentials like Google Chrome, Text Edit, and Finder. Once enabled, ChatGPT begins logging your active work sessions, tracking everything from document edits to saving files and even sending emails.
In practice, it means you can start working without constantly feeding context into the chatbot. Ask “What did I do today?” and ChatGPT provides an overview including key projects, files you handled, and tasks you completed. You can even say, “Create a skill for next time,” and it builds a workflow based on your recorded actions—without extra input.
Context Is Everything for AI
This feature represents a major breakthrough in how AI understands and supports human work. Previously, AI tools relied solely on user input or disconnected memory systems that struggled to grasp the flow and sequence of tasks performed.
Now, ChatGPT sees not just isolated commands but the full timeline of your workday, letting it generate suggestions, automate repetitive tasks, and even identify your most automatable workflows. That’s a level of integration with daily work previously unheard of outside of fictional assistants.
Security and Ethical Concerns Loom Large
While the potential boost to productivity is huge, the implications for privacy and data security are equally profound. Giving ChatGPT broad access to your main work environment means your private communications and sensitive files are potentially exposed to AI systems operated by third parties.
Some users mentioned sandboxing—using separate machines or accounts to test AI tools without risking sensitive data—as a way to mitigate risks. But with ChatGPT’s computer history designed to track your core workflows, careful consideration is required about what you share and what you keep private.
A New Era of AI Assistant Features
This rollout comes at a moment when AI companies race to build more agentic interfaces—that is, AI entities that can act autonomously on your behalf. The combination of computer history and simpler agent-based chat tools might soon make AI assistance feel as natural as talking to a coworker on Slack.
Expect the competition to heat up over the next months between AI firms developing seamless integrations of your work context and conversational agent interfaces. These two features together could finally make AI a genuinely useful partner, rather than only a tool.
Final Thoughts
OpenAI’s computer history feature is a bold step into unprecedented AI integration in daily life, bringing powerful automation possibilities but also raising serious privacy questions. It’s a milestone that highlights how AI’s future will be shaped by both technical capability and ethical boundaries.
Launching this on expensive paid plans and limiting geographic availability shows OpenAI is cautious, but the debate on how much access AI assistants should have is just getting started.
If you value the productivity gains, make sure to carefully configure which apps ChatGPT can see—and pause it during sensitive conversations. This isn’t just another AI feature; it might be the start of the next big shift in human-computer collaboration.
