AI Browsers, a Boon to Efficiency or Bane to Privacy? We Set to Find Out!
Key Takeaways:
- AI browsers collect massive amounts of data: They track your browsing history, downloads, search queries, open tabs, and even sensitive information.
- They can use this data for AI training and ads: The collected data can help them train AI models, build detailed user profiles, and fuel targeted advertising.
- You can still protect your privacy: Use ‘accountless’ modes when available, turn off memory functions, opt out of data training, and avoid sharing sensitive details that you don’t want exposed.
AI companies are coming to your browser.
Perplexity launched its AI browser, Comet, in July this year with much fanfare. OpenAI is also set to launch its own AI browser soon. And Anthropic has already introduced a browser-based agent, Claude, in August 2025.
But behind AI companies’ rush to launch browsers may lie a bigger motive: not user productivity, but user data.
According to a new study by researchers from UCL, UC Davis, and Mediterranean University of Reggio Calabria, AI web browser assistants pose significant privacy risks.
The researchers found that these assistants were collecting and sharing sensitive user data—such as Social Security numbers and medical records—without adequate safety measures.
Read on to find out the privacy risks of AI browsers and how to protect your data while using them.
What Data AI Browsers Collect
After reviewing Comet’s privacy policy, we found that it collects and store the following data:
- URLs of the web pages you visit
- Text, images, and other content from those pages
- Permissions granted to websites (like access to location, camera, and microphone)
- Number of open tabs and windows
- Search queries you type
- Records of files you download from websites
- Cookies stored by the websites you visit
Additionally, it can collect technical data, including your device’s operating system, hardware specifications, and your IP address.
If an AI browser collects this amount of data, I’d be worried. What are they doing with all of it? And what happens if the data is exposed in a data breach?
Perplexity’s privacy notice notes that it can share the collected data with potential buyers, new owners, or other parties during major business changes.
Perplexity CEO Srinivas has been quoted in reputable media platforms, such as TechCrunch and Mashable, stating that he aims to create a browser to track everything in order to sell hyper-personalized ads.
That’s kind of one of the other reasons we wanted to build a browser, is we want to get data even outside the app to better understand you. We plan to use all the context to build a better user profile and, maybe you know, through our discover feed we could show some ads there.
– Srinivas, TBPN Podcast
Perplexity’s CEO even made a $34.5 billion bid to acquire Google Chrome, a bold move from a company valued at $18 billion, nearly half the size of its target.
Why? The answer might be simple: user data. Chrome commands 63.6% of the global browser market. Imagine the volume of user behavioral data Chrome holds.
Another AI-powered browser, Microsoft, launched its AI-based Copilot Mode in the Edge browser, which collects users’ data to improve its product offering.
Microsoft’s privacy policy clearly states that it collects your data to operate, update, and secure its products while offering support when needed.
The company also utilizes your data to enhance services, personalize experiences, advertise, and offer tailored recommendations.
Microsoft being Microsoft, right?
Together, these examples show that AI browsers aren’t just tools for browsing. They’re also extremely efficient data harvesting tools
Why AI Companies Are Rushing Out AI Browsers
AI companies are racing to launch AI browsers primarily to gain control over user data. This is understandable, as AI companies are running out of training data.
According to The Wall Street Journal, AI companies are looking for alternative sources of user data. For instance, OpenAI might use YouTube transcripts to train GPT-5.
AI browsers allows these companies to collect user behavioral data directly, such as searches, clicks, and content consumption habits. They can then use this massive amount of data to train their AI models.
Additionally, AI companies can utilize behavioral data to create user profiles for targeted advertising. Building and running AI models requires a tremendous amount of money, so it’s no surprise that OpenAI remains unprofitable to this day.
Since advertising remains one of the most profitable digital business models, AI browsers offer a strategic way for these AI companies to monetize their technology at scale.
Another reason why AI companies are endeavoring to launch AI browsers is their strategic market position. Launching proprietary AI browsers lets them diversify their product lines and compete with industry leaders like Google.
However, with such deep data comes a real cost: your privacy. AI browsers could easily become surveillance tools, quietly collecting detailed information about your online activities.
Researchers from UCL, UC Davis, and Mediterranean University of Reggio Calabria stated:
Our findings show that GenAI browser assistants collect and share personal and sensitive information for profiling and personalization, highlighting the need for safeguards as they increasingly mediate web browsing.
Safeguarding Your Privacy While Using AI Tools
AI tools, whether chatbots or browsers, collect a massive amount of user data to train the large language models that power them.
As AI companies reach a wall in data training, bringing AI training to users’ browsers is the next step to collect behavioral data.
For instance, Perplexity’s Comet collects previous searches and save memories of previous conversations to better answer your questions and provide a ‘better’ experience.
But what if you don’t want to train the AI with your data or, better yet, keep your browsing activities and preferences private, as Mother Nature intended?
Here are a couple of steps you can take to protect your privacy when using AI browsers:
- Try using the accountless version wherever possible. Doing so can limit the amount of personal information AI tools can collect.
- Avoid using ‘sign in with’ Facebook or Google. Consider using a fake email address to hide your real one. Proton Pass’ hide-my-email aliases are great for this (though a premium feature).
- Delete your data regularly if the tool allows you to do so in the settings menu. Additionally, disable the memory feature if available (Comet has it).
- Opt out of AI training when you can. Google allows you to control the data that Gemini AI has access to on your Android phone.
- Don’t share anything with the AI that you wouldn’t want exposed in a data breach.
Protecting your privacy in the age of AI is a big challenge. These AI tools boost your workflow, research process, and give you a boost in your domain, but often at the cost of your data, privacy and, perhaps, anonymity.
Popular browsers like Edge already have a reputation for tracking users and building profiles for targeted advertising. But AI browsers take user tracking several steps further… much further.
Take proactive steps to protect your privacy when using AI browsers and other AI tools – your data should be yours alone, and you can never have too much privacy.
Sandeep Babu is a cybersecurity writer with over four years of hands-on experience. He has reviewed password managers, VPNs, cloud storage services, antivirus software, and other security tools that people use every day. Read more
He follows a strict testing process—installing each tool on his system and using it extensively for at least seven days before writing about it. His reviews are always based on real-world testing, not assumptions.
Sandeep’s work has appeared on well-known tech platforms like Geekflare, MakeUseOf, Cloudwards, PrivacyJournal, and more.
He holds an MA in English Literature from Jamia Millia Islamia, New Delhi. He has also earned industry-recognized credentials like the Google Cybersecurity Professional Certificate and ISC2’s Certified in Cybersecurity.
When he’s not writing, he’s usually testing security tools or rewatching comedy shows like Cheers, Seinfeld, Still Game, or The Big Bang Theory. Read less
The Tech Report editorial policy is centered on providing helpful, accurate content that offers real value to our readers. We only work with experienced writers who have specific knowledge in the topics they cover, including latest developments in technology, online privacy, cryptocurrencies, software, and more. Our editorial policy ensures that each topic is researched and curated by our in-house editors. We maintain rigorous journalistic standards, and every article is 100% written by real authors.
