Just before lunchtime on Tuesday, Alphabet Inc’s stock price took a sudden 4.8 percent tumble. The bad news? OpenAI was about to announce its long-awaited — some might say late — entry into the artificial intelligence (AI) browser wars.
However, as OpenAI co-founder Sam Altman and his colleagues ran through their browser’s features on a livestream, Alphabet’s stock was retracing its losses. Shares ended the day down 2.2 percent on a flat day for the overall market — perhaps an indication that investors were not as troubled by OpenAI’s play as they might have been.
The product, named ChatGPT Atlas, has bells and whistles powered by its market-leading chatbot. As interesting as they are, they stand little chance of persuading a meaningful number of Chrome users to switch. Consumers might be even more turned off when they learn that the best features of ChatGPT come at a cost. To engage Atlas’ agent — where the AI takes control and handles tasks through the browser — users will need to be US$20-a-month ChatGPT Plus subscribers, Altman said (though he did add “for now”).
This agent aspect is what OpenAI thinks can differentiate the ChatGPT Atlas from other browsers. OpenAI lead researcher Will Ellsworth called it “vibe lifing” — a play on the term “vibe coding,” used to describe software code generated with plain language prompts. In a demonstration, the company showed ChatGPT Atlas analyzing an online recipe and ordering the ingredients by browsing Instacart’s website, searching for the right products and ordering the correct quantity of each.
Neat, but such uses for AI are temporary, like an ’80s kid holding a tape recorder up to a speaker to record songs off the radio. OpenAI clearly understands this, announcing its “Instant Checkout” tool and a partnership with Walmart that allows for ordering without leaving the ChatGPT conversation window at all. That is a far better way to perform tasks than through the slow and essentially manual process of having your AI click around a website, a method that seems inherently vulnerable to breaking or behaving unpredictably. Earlier versions of the same system were described as “glitchy.”
Google’s browser commands a 64 percent share of the US desktop market, rising to 74 percent worldwide, according to data from StatCounter. Chrome is a linchpin of Google’s consumer AI strategy and, thanks to antitrust court decisions, the company can integrate its Gemini bot with abandon. This gives Google a huge opportunity to steer consumers away from typing “chatgpt.com” into their browser in favor of using its own integrations. Microsoft is doing the same with its web browser, Edge (some say unfairly). OpenAI’s lack of a locked-in presence on any computing platform is a growing problem. Use of its mobile app has shown signs of a slowdown, too.
Google’s browser dominance has risen in recent months, even as a number of AI-enhanced browsers have been released, such as the Browser Company’s Dia, Perplexity’s Comet and others. None of these new players are being used enough to appear in StatCounter’s analysis, the company said, all registering below 1 percent share (likely well below). From this we can infer that consumers are not interested. If the browser you have works, and you are used to it, you need a big incentive to change. OpenAI is not offering one.
In many respects, one could argue that ChatGPT’s primary function is to help gather user data to train its AI, bringing on a wealth of extra information on browsing habits that it does not currently have access to. That makes the product release worth it, even if market share stays minuscule.
In trademark fashion, Altman said on Tuesday that AI “represents a rare, once-a-decade opportunity” to upset the browser market. Similar words were said about adding ChatGPT to Microsoft’s Bing more than two years ago — only for Google’s dominance in search to barely move. On Tuesday, Alphabet’s shares took a small hit, yes, but the lack of all-out panic suggests maybe investors are starting to take some of Altman’s hype with a grain of salt.
Dave Lee is Bloomberg Opinion’s US technology columnist. He was previously a correspondent for the Financial Times and BBC News. This column reflects the personal views of the author and does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
We are used to hearing that whenever something happens, it means Taiwan is about to fall to China. Chinese President Xi Jinping (習近平) cannot change the color of his socks without China experts claiming it means an invasion is imminent. So, it is no surprise that what happened in Venezuela over the weekend triggered the knee-jerk reaction of saying that Taiwan is next. That is not an opinion on whether US President Donald Trump was right to remove Venezuelan President Nicolas Maduro the way he did or if it is good for Venezuela and the world. There are other, more qualified
This should be the year in which the democracies, especially those in East Asia, lose their fear of the Chinese Communist Party’s (CCP) “one China principle” plus its nuclear “Cognitive Warfare” coercion strategies, all designed to achieve hegemony without fighting. For 2025, stoking regional and global fear was a major goal for the CCP and its People’s Liberation Army (PLA), following on Mao Zedong’s (毛澤東) Little Red Book admonition, “We must be ruthless to our enemies; we must overpower and annihilate them.” But on Dec. 17, 2025, the Trump Administration demonstrated direct defiance of CCP terror with its record US$11.1 billion arms
The immediate response in Taiwan to the extraction of Venezuelan President Nicolas Maduro by the US over the weekend was to say that it was an example of violence by a major power against a smaller nation and that, as such, it gave Chinese President Xi Jinping (習近平) carte blanche to invade Taiwan. That assessment is vastly oversimplistic and, on more sober reflection, likely incorrect. Generally speaking, there are three basic interpretations from commentators in Taiwan. The first is that the US is no longer interested in what is happening beyond its own backyard, and no longer preoccupied with regions in other
As technological change sweeps across the world, the focus of education has undergone an inevitable shift toward artificial intelligence (AI) and digital learning. However, the HundrED Global Collection 2026 report has a message that Taiwanese society and education policymakers would do well to reflect on. In the age of AI, the scarcest resource in education is not advanced computing power, but people; and the most urgent global educational crisis is not technological backwardness, but teacher well-being and retention. Covering 52 countries, the report from HundrED, a Finnish nonprofit that reviews and compiles innovative solutions in education from around the world, highlights a
RSS