Atlas and Google Chrome browsers are recognized as the most insecure in the world.

ccording to the results, OpenAI's ChatGPT Atlas was recognized as the worst browser. It received the highest risk level—99 points out of 100. The browser failed key tests, particularly state separation, meaning it lacks protection against user activity tracking across sessions. Low scores were given for all three parameters examined: privacy and fingerprinting protection, tracker and data blocking, and connection security.

The next least secure browsers were Google Chrome (76 points), Vivaldi (75 points), Microsoft Edge (63 points), and Opera (58 points). Higher security scores were demonstrated by Ungoogled (55 points), Mozilla Firefox (50 points), Apple Safari (49 points), DuckDuckGo (44 points), and Tor (40 points).

Digitain's Paruyr Harutyunyan noted that AI browsers like ChatGPT Atlas and Perplexity's Comet are attracting attention and millions of users due to the hype surrounding AI.

"These browsers have interesting features that traditional browsers don't, but users should stop and check how secure they actually are. AI works by collecting and learning from data, which means these tools may be collecting more of your personal information than you realize. Just because something uses AI doesn't automatically make it secure or private."

Paruyr Harutyunyan, Head of Digital Marketing at Digitain

Brave and Mullvad Browser emerged as the most secure browsers in the study. The latter is an open-source browser focused on maximum privacy and minimizing user data collection.