Also according to official material released by the company, 75% of accounts removed for inappropriate actions are identified through the app's machine learning algorithm. To learn more about how this intelligence system works, check out the strategies developed by WhatsApp software engineering.
Open WhatsApp groups can be dangerous; understand the risks
Four tips to make you invisible on WhatsApp
Want to buy a cell phone, TV and other discounted products? Meet the Compare dnetc
The Three Stages of Spam Fighting
To ban users without reading the content of private conversations, WhatsApp uses in-app log metadata and message submission rate. Through this information, you do not need to decrypt the chat. Everything happens through the system of Machine learningis a computer science field that combines artificial intelligence and model recognition. In this way, abuse detection takes place in three stages. First, at the moment of account registration. Then while writing and sending messages. Finally, negative reactions, through denunciations and blocking, also help in combating abusive accounts.
At the time of registration, the device data where the created account allows WhatsApp to verify its coordinates. The phone number, mobile phone IP address and other details can be used to detect the source of a problem. The messenger can verify that the account is being created from the same phone or network that has shown suspicious activity. In this case, the application is likely to remove the account when the user attempts to register it. Just in the last three months, WhatsApp has revealed that 20% of accounts were banned at the time they were registered.
Real-time message evaluation is the second stage. Here, the intensity of use that makes the difference. It is relatively easy to identify abuse at this point after all, well-meaning users use the application sparingly when forwarding content only occasionally. On the other hand, if a person sends 400 messages per minute, for example, it is very likely that the account is related to a messenger misuse scheme.
Understand how WhatsApp fights spam without reading encrypted messages Photo: Paulo Alves / dnetc
Negative reactions to an account are, finally, another tool that allows users to be banned without messages being read. Sent reports are categorized by WhatsApp and allow you to understand the motives of accounts sending unwanted messages, such as spreading false information (fake news) or even sell a product. The software also stimulates protection against sending improper messages by offering the option of blocking information to unknown numbers.
To improve the identification of negative reactions, the app also filters out unfair blocks. Thus, just as WhatsApp strives to identify excessive messaging, a large number of reports directed to the same user can also be investigated. One way to do this is to verify that the phone numbers that made the reports actually interacted with the person you reported.
Apparently the measurements are working. In February 2018, the company reported that it has 1.5 billion active users per month. Of this total, about two million accounts are removed monthly. About 25% of users are excluded due to complaints from others. Most banned, therefore, thanks to the software engineering efforts that WhatsApp has been promoting.
Other initiatives against misuse
In addition to fighting automated messaging, there are other ways to identify abusive behaviors. One of them is the message forwarding warning. With the feature, the user can find out if the message received was actually created by the sender.
Since last month, WhatsApp has limited message forwarding to only five contacts. In response to dnetc, the company says the function "will help keep the app focused on private messages with close contacts." Initially, the app allowed the same content to be sent to more than 200 people simultaneously.
The label for "Encaminhada" comes in shared messages Photo: dnetc / Helito Bijora
Challenges that persist in fighting spam
One difficulty the company faces are unauthorized modified versions of the application such as GB WhatsApp and Yo WhatsApp. The programs violate the company's terms of service and endanger the safety of users. This is why the best protection is to download WhatsApp only from the official website or from the Android and iPhone (iOS) app stores, ie Google Play and App Store, respectively.
Another challenge is the demand for cracking encryption. Although the company publicly states that it does not have access to user messages, there are cases where this request was made by authorities in criminal cases.
Finally, the biggest challenge facing WhatsApp may be the modifications made to the devices themselves where the app is used. In a meeting with journalists in India, one of the software engineers, Matt Jones, reported that many abusive users use altered cell phones with multiple SIMs, or simulators that can perform multiple WhatsApp inputs.
Another problematic practice when abusive users create new groups on the same topic and then repeatedly add the same people to the conversations. This activity diminishes the effectiveness of the action of denouncing and leaving a group.
WhatsApp: Five Tips for Using the App Safely
And the user, what can you do in this fight?
Despite all the advances, it is clear that much remains to be done. It is always worth remembering that, in addition to the company's efforts, users themselves can also make important decisions. A decisive act in not sharing rumors or unfounded news is how to identify fake news on WhatsApp. In countries such as India or Brazil, where election periods have been affected by group or automated messaging, this is an act of citizenship. Also, reporting accounts that make inappropriate use of WhatsApp can also make a lot of difference to the app's user community.