AI browsers bypass paywalls on websites without any auxiliary tools.
New AI browsers, such as Atlas by OpenAI and Comet by Perplexity, easily bypass paywalls* – pages on websites that are accessible to readers only with paid access. However, it is important to understand that the purpose of these technologies is not to infringe copyright, but to make knowledge more accessible and searchable.
*What is a pavilion:
Paywall is a system that limits access to content, allowing you to read only a part of an article or several materials for free. This is how publications finance their work, support journalism, and develop quality content.
Atlas and Comet read paid articles as if it were a regular Chrome browser, while ChatGPT just says sadly: “No, I won’t let you.” They know how to bypass paywalls without any tricks, while other bots get a digital red flag and are blocked. Who would have thought that being a “human” for a website is so easy?
No “hacker skills” are required to do this – just ask the AI browser to show the full text of the article. In an experiment by Columbia Journalism Review, Atlas and Comet had no trouble publishing an exclusive 9,000-word MIT Technology Review article, while standard chatbots like ChatGPT and Perplexity honestly reported: “Access blocked”.
The secret is simple: for Atlas and Comet, websites are ordinary Chrome users, not suspicious automated scanners with digital badges. The latter, in turn, are easily banned through the Robots Exclusion protocol, a verbal red flag for bots.
At MIT Technology Review, the text is hidden behind a banner asking to subscribe, but the content itself is already uploaded to the page – invisible to human eyes, but fully accessible to AIs like Atlas and Comet. The Wall Street Journal and Bloomberg are even stricter: the text appears only after login and payment, so even the smartest AI without keys won’t get through.
AI can spy on all the banners, but a real pavilion reminds you: “Even the smartest bot will have to pony up for this one.”
