AI bypasses paywalls of news sites: study reveals details.

Test on bypassing paywalls

According to inkorr.com: Artificial intelligence, as well as OpenAI and Perplexity browsers, successfully passed the test on bypassing paywalls on news resources. Such results were published by the Columbia Journalism Review, which showed that bots can access content intended only for paid subscribers without any problems. Thus, they are effectively stealing information.

Testing Results

During the test, the Atlas browser from OpenAI and Comet from Perplexity easily accessed the full text of an MIT Technology Review article that contained 9,000 words and was only available by paid subscription. In contrast, other browsers, including ChatGPT and Perplexity, faced site-level blocking and could not access the material.

The secret of success lies in the fact that AI-based browsers appear to websites as ordinary users using Chrome. They leave no traces of their true nature, unlike traditional bots.

Content Security Issues

Web scrapers and parsers usually identify themselves through a special digital marker that can be blocked by publishers using the Robots Exclusion Protocol.

All these facts raise new questions about content protection on websites and about the further use of artificial intelligence in research. The situation indicates that traditional protection methods may prove insufficient, and there is a need to improve information protection technologies in the context of rapid artificial intelligence development.