The Browser Wars Are Back With AI Agents

News

1. India Faces Lower Risk of AI Job Disruption Than the West, Says IT Secretary

2. Tata Steel Discloses $1.6 Billion Dutch Class Action Over Emissions

3. Nvidia Licenses Groq’s Inference Tech, Hires Leadership

4. HCLTech Deepens Software Push With Three Acquisitions in a Week

5. Isro Launches BlueBird Block 2 in Heaviest Commercial Mission Yet

6. OpenAI Softens ChatGPT’s Tone While Scaling for an AI Showdown

For most users, the web browser has been a solved problem for years. Chrome dominated, Edge lingered, Firefox survived on principle, and few people thought twice about what sat between them and the internet.

In 2025, that complacency broke. Across Silicon Valley and beyond, Big Tech firms and AI startups began treating the web browser as the most valuable real estate on the internet. 

The reason? 

Whoever controls the browser increasingly controls how users search, decide, shop, work, and feed data back into AI systems.

Market research firm Market.us estimates the global AI browser market could grow from $4.5 billion in 2024 to $76.8 billion by 2034, reflecting how quickly companies believe this layer of the internet is being reshaped. North America already accounts for more than a third of current revenue.

What changed this year was technology, and also intent. Browsers stopped being passive windows and started acting like assistants, agents, and, in some cases, decision-makers.

 

Microsoft makes the browser an enterprise tool again

Microsoft Corp. was the most pragmatic mover.

Rather than launch a separate product, it turned Edge into an AI-assisted work surface by embedding Copilot directly into the browser. Copilot Mode analyses open tabs, summarizes pages, drafts emails, compares information, and performs simple actions like bookings or unsubscribing from newsletters.

Microsoft’s pitch was aimed squarely at enterprises. By integrating Copilot with Microsoft Graph, Office apps and existing security controls, the company made Edge an AI tool that corporate IT teams were more willing to tolerate than standalone AI software.

There were early reliability issues, particularly with Copilot Actions, but updates through October improved performance. Edge remains pre-installed on Windows machines, giving Microsoft distribution that rivals cannot match.

 

Opera bets on automation over chat

Opera took a more experimental path In May, it relaunched Opera Neon as an “agentic browser”, designed not just to answer questions but to carry out multi-step tasks. Neon builds on Opera’s earlier Aria assistant, which already handled contextual prompts, translations, and image generation.

The difference is posture. Aria reacts. Neon acts.

Users can ask Neon to plan research, organize tabs, or manage workflows, with the browser behaving more like a lightweight operating system. Neon is a paid product, while Aria remains free and widely used.

Opera’s global market share is small, or around 2–3%, but it punches above its weight in gaming and mobile through Opera GX. Growth in 2025 was real but constrained by one thing Opera has always struggled with: visibility.

 

Perplexity goes straight for Google

Perplexity’s Comet browser was the most openly confrontational. Rather than layering AI onto browsing, Comet turns the address bar itself into a conversational agent. Users ask questions, run comparisons, summarise emails, or fill forms without ever thinking in terms of “search”.

By late 2025, Perplexity said Comet had reached 12 million monthly active users, driven by researchers and power users frustrated with traditional search results.

Comet is fast and well-cited, but it runs into a familiar wall: enterprise trust. Many organizations remain cautious about letting AI-native browsers operate freely inside corporate environments, especially without long-standing security frameworks.

 

OpenAI puts ChatGPT at the centre of browsing

In October, OpenAI launched Atlas, its own browser built around ChatGPT. Atlas defaults to conversation rather than search. New tabs open into ChatGPT. An agent mode allows supervised automation across tabs, including shopping and research. The idea, according to CEO Sam Altman, was to rethink browsing around intent rather than clicks.

Early feedback has been mixed. Users like the interface and contextual awareness, but some report slower execution on complex tasks and occasional bugs. Still, Atlas saw unusually fast early uptake, particularly among developers and AI-native users.

Atlas is free at the base level, with advanced agent features tied to paid ChatGPT plans.

 

Google defends its core advantage

Google, unsurprisingly, played defense and offence at once. Gemini is now embedded into Chrome, offering contextual assistance across tabs, deep recall of browsing history, and tight integration with Google services such as Maps, Calendar, and YouTube. New Gemini models introduced this year brought faster responses, better reasoning, and multimodal capabilities.

 

Alongside Chrome, Google quietly launched Disco, an experimental browser that auto-generates small web applications based on user intent. Disco’s GenTabs feature builds task-specific tools without code, always linking outputs back to source material.

Google’s advantage remains scale. Chrome already owns the browser market, and Gemini benefits from Google’s data, distribution, and advertising machine. The risk for Google is not adoption, but disruption: AI browsers threaten the very search-driven revenue model Chrome was built to serve.

 

Mozilla chooses a slower lane

Mozilla did not launch a new browser in 2025, but it did reposition Firefox.The company introduced AI features such as the AI Window, an opt-in assistant designed around user choice and privacy. 

New CEO Anthony Enzor-DeMeo said Firefox would evolve into a “modern AI browser” supporting multiple models, including open-source ones, without forcing users into a single ecosystem.

Firefox’s approach is deliberately cautious. It will not win speed races, but it continues to attract users wary of surveillance-driven AI.

 

What this was really about

This was never just about browsers.

In 2025, companies realized that AI needs a home, and the browser is the most natural one. It sits where intent is expressed, decisions are made, and data is generated. Control that layer, and you control far more than traffic.

No single player won the browser war this year. But the old idea of browsing seems to be gone.

The browser is today where intelligence lives, watches, and increasingly, acts. That is the shift 2025 locked in.