IT News Correspondent
By Juliet Orbo
OpenAI has officially entered the web browser arena with the launch of Atlas, a revolutionary new browser powered by its flagship artificial intelligence model, ChatGPT. The company says Atlas is designed to challenge the dominance of Google Chrome and “reimagine what a browser can be.”
Announced on Tuesday, the ChatGPT-driven browser aims to go beyond traditional browsing by turning internet navigation into a conversational, intelligent experience. According to OpenAI CEO Sam Altman, Atlas represents “a rare, once-in-a-decade opportunity to rethink what a browser can be about and how we use one.”
A Browser Without the Search Bar
Unlike conventional browsers that rely on URL bars and manual search, Atlas introduces a chat-based interface that allows users to simply “ask” the browser for what they need. Instead of typing links or keywords, users can make natural language requests — from booking a hotel and summarizing research papers to comparing prices — all without switching tabs or copy-pasting data.
“Tabs were great, but it’s time to move forward,” Altman said during a virtual demo. “Atlas helps you complete tasks directly in the browser without constantly jumping between websites or applications.”
Initially, Atlas will be available only for Mac computers, with a Windows version expected to follow later. OpenAI promises seamless integration with ChatGPT, allowing users to browse, analyze, and act — all from a single interface.
Agent Mode: The Browser That Works for You
One of Atlas’s standout features is Agent Mode, which allows the AI to browse the internet automatically on the user’s behalf. By analyzing browsing habits and previous searches, the browser predicts what the user might need next, retrieving relevant data proactively.
In essence, it doesn’t just find information — it interprets and acts on it. For example, if you’re researching flights, Atlas could compare prices, suggest nearby hotels, or even draft an email to confirm a booking.
“It’s using the internet for you,” Altman explained. “Instead of you hunting for answers, Atlas hunts for them — and delivers insights in plain language.”
Privacy and Personality Concerns
While Atlas’s capabilities are groundbreaking, experts are already voicing privacy and autonomy concerns. Paddy Harrington, a senior analyst at London-based Forrester Research, cautioned that AI browsers like Atlas could blur the line between personalization and control.
“Atlas might feel like it knows you — maybe too well,” Harrington told the Associated Press. “It’s collecting vast amounts of data to predict your behavior. But at what point does the browser stop reflecting your personality and start imposing its own?”
Harrington also warned that algorithmic personalization could open the door to subtle advertising bias. “If the AI decides which sources to trust or which products to recommend, who’s really in control — you, or the engine behind it?”
A Giant to Challenge
Atlas is entering a market long dominated by Google Chrome, which has held the top spot since its launch in 2008 and currently boasts an estimated 3 billion users worldwide. Competing with that scale is no small feat.
Still, OpenAI may have an edge. As generative AI tools become mainstream, more users are turning to ChatGPT and similar models for quick, summarized answers — bypassing traditional search engines altogether. OpenAI reports that ChatGPT has surpassed 800 million active users, and a 2025 survey by the Associated Press found that 60% of Americans, and 74% of those under 30, now use AI tools to find information at least part of the time.
Are AI Browsers Trustworthy?
Despite their growing popularity, AI-powered browsers face one critical challenge: accuracy. A new study released by the European Broadcasting Union (EBU) and the BBC revealed that 45% of AI-generated answers contained at least one major factual error, while 81% included some form of misinformation, bias, or misinterpretation.
The study, which evaluated models such as ChatGPT, Gemini, Copilot, and Perplexity in 14 languages, found that AI systems often struggle to differentiate between fact and opinion. In one example, ChatGPT confidently stated that the late Pope Benedict XVI was still alive — a seemingly small but serious error in factual integrity.
“When people can’t trust what they read, they start trusting nothing,” said Jean Philip De Tender, the EBU’s Media Director. “That uncertainty doesn’t just undermine technology — it undermines democracy.”
The Future of Browsing: Assistance or Automation?
As Atlas rolls out to early adopters, the world will be watching to see whether it truly revolutionizes browsing or simply deepens our dependence on AI intermediaries. While some hail it as a bold step toward a smarter, more intuitive web, others see it as a warning sign — one that challenges the very nature of personal agency and information autonomy.
OpenAI insists Atlas is built with transparency and user control at its core. Yet the question remains: can a browser that “thinks for you” also be one that lets you think freely?
For now, one thing is clear — the race to define the next era of web browsing has officially begun, and with Atlas, OpenAI is betting that conversation, not clicks, will shape the future of the internet.
