12 September 2021 – ° BLOG ° Gabriele Romano

Conversations on WhatsApp are protected by end-to-end encryption, as are most of the shared items. However, according to a new long article written by ProPublica, not all messages exchanged on the client, now owned by Facebook, are safe from prying eyes. Facebook or third-party consultants hired by the company, in fact, can read some extracts of the conversations that take place on the platform.

The heavy accusation is brought by the source through a long article that reads how Facebook has been sponsoring for years the use of end-to-end encryption on WhatsApp as a fundamental tool to keep any data exchange that takes place on the service safe. Despite this, the company can gain access to certain parts of the conversation, and ProPublica can say this based on observations made of about a thousand firm’s contract consultants who have reviewed millions of data.
WhatsApp, the accusation: contractors paid to read parts of conversations

These contractors have access to a dedicated Facebook software that is used to monitor private messages on WhatsApp, images and videos, which have been reported by other users: “These contractors have the task of analyzing everything that is displayed on their screen – statements about anything like fraud, spam or child pornography, or even potential terrorist plots – in less than a minute, “an excerpt from the report reads.

It should be noted that the contractors do not have the possibility to read the conversations of the users in full, but only the last five lines of a one-to-one conversation that has been reported by one of the two participants. This modus operandi is also indicated on the client itself when trying to report a conversation as inappropriate. The workers contracted by Facebook to monitor WhatsApp conversations operate in Austin, Dublin, Singapore and Texas, and carry out the monitoring activity as the sole task for the giant of Menlo Park.

If on the one hand it is true that consultants paid by Facebook can access parts of the conversations, on the other hand it is also true that these are sensitive chats reported as inappropriate and potentially dangerous by users. This could clash with the “pro-privacy” narrative spread by Facebook regarding the fact that the company cannot in any way access the contents of the chats due to end-to-end encryption, but in this case the purpose is to make secure conversations for all users.

Carl Woog, communications director for WhatsApp, admitted that the different contractors also have the ability to identify and remove the “worst” abusers. The same monitoring is also done on Facebook and Instagram, since Artificial Intelligence alone cannot be 100% effective to protect the most defenseless users from inappropriate, dangerous or even annoying content.

Facebook promptly responded to the accusations made by ProPublica, and below is an excerpt from the note released:

"We have developed WhatsApp to limit the data we collect to obtain spam and threat prevention tools, thus being able to remove abusive users, even based on the reports we receive from users. This work requires extraordinary effort. from our security experts and the trusted teams we employ, who work tirelessly to help us provide the world with secure private communication. "

The company also specified that “based on the feedback received from users, we are sure that people understand that when a conversation is reported to us, we receive an extract from it to be able to analyze it”.