Many right-wing news sites allow AI companies to scrape their data

Hotstar in UAE
Hotstar in UAE

The intersection between AI and politics is getting a lot more foot traffic nowadays because of the impending U.S. election. Not only that but there are other elections happening around the world this year. Well, a new report uncovered the fact that many right-wing publications allow their content to train AI chatbots.

In order to train AI chatbots, companies need to send out crawlers to different sites. crawlers travel to the sites and scrape text data. This is something that gets certain companies in hot water, as the crawlers could potentially scrape copyrighted information. Currently, Microsoft and OpenAI are in the midst of several lawsuits. The New York Times claims that the companies use copyrighted articles to train their AI models. Also, a handful of other publications including New York Daily News is suing the companies for much the same reason. This is why most publications choose to block these crawlers.

Most right-wing publications allow their data to train AI chatbots

Recent data reviews that about 88% of top news publications do not want crawlers on their sites. So, it’s pretty obvious where most publications stand on this matter. However, Ontario-based AI detection startup Originality AI conducted a study on 44 top news sites. The study revealed that nearly all of the sites block crawlers as well. However, of the sites, nine of them were right-leaning publications. All of the right leading publications opt to have crawlers scrape their data. These publications include Fox News, Breitbart, and the Daily Caller.

Obviously, studies like these should not be taken as objective proof of a party’s agenda. However, the fact that all of the surveyed right-wing publications allow for crawlers does raise some speculation. Some sources have argued that this could be a tactic to steer AI toward right-wing ideals.

The thing is that AI chatbots are more likely to reproduce information based on the information that they’re fed. So, if a chatbot is predominantly receiving information from right-leaning publications, then it could ostensibly lie right itself. However, that is yet to be seen

2024-05-01 15:05:52