Microsoft doesn’t plan on rolling out a brand-new AI browser; instead, its AI CEO Mustafa Suleyman says the company will evolve its existing Edge browser by embedding Copilot deeply enough that the AI can control tabs, read pages, and perform browsing tasks while you watch and intervene. In his interview, Suleyman describes a vision of a “true agentic browser” where your AI assistant acts like a partner in exploration rather than a passive tool. He positions this strategy as more practical than building a fresh browser from scratch, emphasizing transparency, user trust, and maintaining traffic to publishers. Meanwhile, AI browser competition is accelerating: Perplexity has officially launched its Comet browser for Pro users in India (and beyond), built on Chromium with AI agent features; OpenAI is reportedly preparing a browser codenamed “Aura” to challenge Chrome; and Google is weaving its Gemini AI more deeply into Chrome itself. But Comet has not escaped scrutiny—independent audits from Brave and Guardio have flagged vulnerabilities like prompt injection and malicious code risks.
Sources: Reuters, Windows Central
Key Takeaways
– Microsoft’s strategy is evolutionary, not revolutionary: instead of launching an entirely new browser, it aims to upgrade Edge by integrating Copilot as a task-performing agent.
– The competition is intensifying—Perplexity’s Comet is live for paying users, and OpenAI, Google, and others are pushing AI features into browsers or building new ones.
– Security is a major concern: Comet has been flagged for vulnerabilities such as prompt injection and exposing user data, underscoring the risks when AI takes control of browsing.
In-Depth
The browser — that long-trusted workhorse for web navigation — is under pressure to evolve. Microsoft’s AI division head, Mustafa Suleyman, recently laid out how the company sees the future: not by throwing away Edge and building something entirely new, but by transforming Edge into what he calls an “agentic browser,” one in which Copilot can actively manage your tabs, surf the web, read, compare, synthesize, and assist in real time while you watch. He frames this as a blend of AI utility and human oversight: you’ll always be able to step in or turn the AI features off if you prefer the classic browsing mode.
This direction seems crafted to preserve Microsoft’s existing user base and relationships with publishers, because COPILOT actions still count as regular browser traffic, keeping ad and content ecosystems intact. Suleyman contends that rather than trying to out-innovate rivals by creating a brand-new browser, it’s more sustainable to layer AI capabilities into what people already use.
At the same time, Microsoft is not alone. Perplexity, a rising star in AI search and agents, has launched the Comet browser—initially for Pro (and earlier Max) subscribers—in India and other markets. Built atop Chromium, Comet brings AI features like summarization, sidebar assistance, task automation, and deep integration with Perplexity’s search engine. Meanwhile, rumors and leaks suggest OpenAI is building its own browser (reportedly named “Aura”) that embeds ChatGPT functions and agentic action, while Google is weaving Gemini capabilities into Chrome’s interface, making Chrome a more active AI browser.
But the shift toward AI-driven browsing isn’t without peril. Security researchers have already raised alarms about Comet’s vulnerabilities. Brave and Guardio audits revealed that how Comet handles webpage content leaves openings for prompt injection, malicious command execution, and phishing attacks—weaknesses that arise when AI is allowed to act rather than simply provide suggestions. These hazards highlight the core tension of allowing an AI to control browsing: convenience and automation come at the cost of giving up some control, and potential attack surfaces multiply.
The stakes are high. Whoever succeeds in making AI browsing seamless, trustworthy, and secure may reshape how we navigate the web, break or remake browser dominance, and drive what expectations users have of intelligent assistants. Microsoft’s gamble is that evolution — rather than revolution — gives it a competitive edge. But others are betting on clean-slate AI browsing models. In this fast-moving race, security, transparency, and user control may ultimately decide who wins.