The developers of the Guard service using artificial intelligence analyzed the user agreements of the largest social platforms and instant messengers, with which users often agree without looking, Mayak writes.
A summary of the texts of the agreements Instagram, Twitter, YouTube, Netflix, Tinder, Aliexpress, Shazam and Telegram is now available on the site.
Based on the results of this analysis, the AI determined that Instagram and Twitter are the most unsafe for user data.
The most dangerous clause in the Instagram agreement, according to the algorithm, looks like this:
“Instagram cannot guarantee the security of any information you submit to Instagram, or guarantee that the information will not be accessible, disclosed, altered or destroyed.”
Twitter has this phrase:
“Your information may be sold or transferred as part of this transaction.”
The safest from the point of view of the Guard is the Telegram messenger. Algorithm concerns raised only this clause of the agreement:
"We will never delete your pictures with funny cats, because we ourselves are delighted with them."
It is noteworthy that the site focused on popular services for some reason did not appreciate the social network Facebook. However, developers accept proposals for resource analysis by filling out a special form.
They also offer users to train artificial intelligence in a game form to better recognize unacceptable points in agreements - you need to choose one of two options for the wording in the document.
Recall that the social network Facebook has repeatedly been implicated in scandals with the leak of user data. And in August, company management admitted that they hired hundreds of third-party contractors to listen to the audio recordings they send in the Messenger app.