Microsoft has rolled out a new feature in its Edge browser dubbed Copilot Mode, a move that pushes the company further into the growing competition around AI-powered web browsing tools.

Unveiled on Monday, the feature turns Edge into an intelligent assistant capable of not just answering questions but actively helping users research, compare, and take action online, from booking appointments to navigating multiple tabs of information.

The launch comes amid a surge in demand for more intuitive and productive browsing experiences powered by artificial intelligence. Microsoft is positioning Copilot Mode as a key leap forward, integrating AI into the browser environment in a way that removes friction and anticipates users’ needs.

Copilot Mode, still in experimental rollout, is opt-in and currently free for all Edge users on Mac or PC who already have access to Copilot. Once enabled, users are greeted with a revamped new tab experience where they can search, chat, and navigate the internet with the AI’s assistance. On any web page, the AI can help contextualize what users are reading or offer actions based on their browsing intent.

One demonstration showed how Copilot could help convert a standard recipe into a vegan version, suggesting plant-based substitutions without the user needing to rephrase or copy content into a separate chatbot. In another example, the AI companion can simply extract and present the essential parts of a webpage — such as the core ingredients of a recipe — skipping over the typical long preambles that often bog down web content.

More than just a chatbot, Copilot is designed to act like a research assistant or task manager. It can draft content, generate shopping lists, and even assist in booking hotels or flights — a function that blends search with intelligent filtering. While this kind of “agentic” behavior signals a big step in browser evolution, Microsoft acknowledges that the experience may not yet be faster or more intuitive than manual navigation, especially for seasoned web users.

Notably, Copilot allows for voice input, opening up accessibility for users who are less comfortable with digital navigation or who may have physical limitations. The feature is expected to evolve to handle more complex tasks as users permit access to browsing history, credentials, and additional context, but for now, much of the action is still manual.

Where Copilot may shine most is in research-heavy sessions. With the user’s permission, the AI can access all open browser tabs to understand the context and patterns in browsing activity. For instance, if someone is comparing flight prices across different websites, Copilot can synthesize the options and present a summary or suggest next steps. Eventually, the tool will be able to recommend where users left off on a project or nudge them forward with suggestions based on their activity.

Microsoft emphasizes that privacy controls are central to Copilot Mode. The AI can only access browsing data when explicitly authorized by the user, and this access will be visibly flagged with clear indicators in the interface. Still, the notion of an AI assistant that can “see and hear” what users are doing online — even if permission-based — raises concerns around digital surveillance, especially at a time when tech giants are under scrutiny for how they manage data.

The Race Toward AI-First Browsing

Microsoft’s move comes amid a broader race to redefine how people interact with the internet using AI. Browsers are rapidly evolving from passive information portals into active, conversational tools that aim to reduce cognitive load and take action on behalf of users. With companies like Google also integrating AI into Chrome and startups pushing AI-native browsers from the ground up, Microsoft is banking on Copilot Mode to keep Edge relevant and competitive.

The browser, once a passive gateway to the web, is now becoming a full-fledged digital assistant. However, some analysts note that the success of Copilot Mode will ultimately depend on whether users find it truly helpful or more of a gimmick than a necessity.

As artificial intelligence reshapes the modern workplace and transforms how people write, code, and create, Bluesky CEO Jay Graber is warning that one thing must remain intact: critical thinking.

According to her, the temptation to hand over reasoning tasks to machines could weaken our most essential human skill—judgment.

“AI is able to automate a lot of critical-reasoning tasks, and if we fully outsource our own reasoning, it’s actually not good enough to run in an automated fashion,” Graber told Business Insider in a recent interview. “You can’t just fully outsource your thinking, or an essay, to AI.”

Graber’s comments come at a time when AI tools like ChatGPT and Claude are increasingly used for essay writing, customer support, moderation, and even software development. Bluesky, the decentralized social media platform she leads, uses AI to assist with moderation and content curation. But Graber stressed that Bluesky never allows AI systems to operate autonomously. Every AI-generated suggestion is reviewed by a human because context, she said, is everything.

“When you let it run autonomously, it doesn’t have actual context or intelligence, or the many things that we need as humans to make good decisions,” she explained. “And so it’s producing stuff that sounds or looks right without actually being right.”

AI Can Do the Work—But Do You Know How to Judge It?

Graber emphasized that as AI becomes more capable, the human role is evolving—from direct output to judgment and refinement. Whether it’s writing or coding, she believes users must still understand the fundamentals.

“If you don’t know what good code looks like, if you don’t know how to actually build a system, you’re not going to be able to evaluate its output,” she said.

AI can generate text or even solve bugs, but without the ability to assess and correct what it produces, the risk of flawed reasoning or misleading content grows.

Build the Thinking Muscle—Don’t Let It Atrophy

Graber advised students to push back against convenience and actually write essays by hand. The goal, she explained, is to build up the “muscle for critical thinking”—not weaken it by overusing AI shortcuts. The long-term risk, she warned, is a world filled with content that “sounds or looks right without actually being right.”

She also championed a generalist mindset, encouraging individuals to develop a wide range of skills rather than relying on AI as a crutch. In her view, AI is like packaged expertise, but human discernment is what gives that knowledge value.

“You need to have the good judgment of how you’re going to use it, and then you have to have the flexibility to take that knowledge and do something useful with it,” Graber said.

The Takeaway: Use AI, But Don’t Let It Use You

As artificial intelligence becomes a core part of everything from business to education, experts like Graber and Chowdhury are urging caution, not because AI is inherently harmful, but because human judgment is irreplaceable.

Bluesky’s refusal to let AI run unchecked is a case in point. “We’re never going to let it make decisions on its own,” Graber said. “Because it doesn’t understand what’s at stake.”

In sum, Graber is saying: in a world increasingly shaped by algorithms, the ability to think critically, ask the right questions, and apply sound judgment may be the most valuable skill of all.