Enhanced data protection on WhatsApp: Fact check on chain letters

The message sounds alarming: "As of today, AI is available on WhatsApp and has access to all chats." To prevent this, the "Enhanced Privacy" option must be activated immediately, it continues – otherwise, artificial intelligence could access group messages, see phone numbers, and "even retrieve personal information" from the smartphone.

This even applies to private chats, the message states. It ends with an urgent plea: people should change their settings "as soon as possible" and share the message "with other groups you belong to."

Digital chain letters like this one have been circulating thousands of times on WhatsApp for several weeks now – and causing confusion. Not entirely without reason: The new "Enhanced Chat Privacy" feature actually exists – and it also has something to do with artificial intelligence. However, it doesn't do exactly what the chain letter suggests.

The so-called "Enhanced Chat Privacy" was announced in the spring of this year. Parent company Meta published a blog post explaining the new feature. The goal: Content shared on WhatsApp with friends, family, and acquaintances should remain there and not be easily shared.

Specifically, the feature does the following: When enabled, it blocks other people from easily exporting charts. Furthermore, media is no longer automatically downloaded to the smartphone's photo gallery. However, the feature can't completely prevent this. For example, it's still possible to take a screenshot of a chat.

In fact, there's also a change regarding artificial intelligence. On WhatsApp, it's possible to ask the Meta AI questions about chats by mentioning it in a chat with "@Meta AI" – then the respective message is sent to the AI. You can also have the Meta AI summarize messages. In these cases, Meta's servers inevitably gain access to the respective message to provide a suitable response. This can now be disabled with "enhanced chat privacy."

What's not true: The AI ​​isn't inherently connected to chats. WhatsApp messages are end-to-end encrypted and—contrary to what the chain letter says—cannot be read by Meta by default.

Denmark has held the EU Council Presidency since July. The country is a strong supporter of so-called chat control, the government's scrutiny of private messenger messages. Whether it actually comes to fruition could now depend on the German government.

Anyone who still has concerns can at least feel better about the "enhanced chat privacy" feature. With this feature, Mark Zuckerberg's company appears to be responding to criticism from its users. In recent weeks, social media has repeatedly expressed discontent about the fact that the meta-AI can simply be "invited" into chats.

It can be deactivated in both private and group chats. Contrary to what some chain letters claim, group members without administrative rights can also do this. Everyone else will then see a notification in the chat that the protection has been activated. However, a group admin can specify that this setting can only be made by admins. This does not, however, completely deactivate the meta-AI. It is still available on WhatsApp as a dedicated chatbot via the blue circle.

Enabling "enhanced privacy" doesn't necessarily mean any disadvantages. However, if you explicitly want photos to be automatically saved to your device's gallery, you won't be happy with the feature—it's simply no longer possible.

If you want to activate the “extended data protection” function, you can do so as follows:

• On Android devices, click the three dots in the respective chat and then click "View contact." In the field further down, you can activate the "Enhanced chat privacy" button.
• On iOS devices, you can access this feature by clicking on the contact in the chat, which will display the contact information. The "Enhanced Chat Privacy" feature is also located in the field further down.

In group chats, the feature can be activated under “Group Info” and then under “Group Permissions.”

Even though the circulating chain letters sound alarmist and sometimes spread misinformation, they nevertheless have a kernel of truth. In March, WhatsApp introduced AI features in WhatsApp in the EU. In May, the company began training its AI using data from EU users. However, the messenger app WhatsApp is excluded from this process – it instead concerns public posts on the Instagram and Facebook platforms. Anyone who does not want their data used can object using a form . The deadline for past posts has since expired.

Another piece of news made the rounds in June: It was revealed that Meta and the Russian-Dutch web company Yandex had allegedly used extremely dirty tricks to spy on Android users in order to create advertising profiles. According to a team of scientists, they secretly established a data connection between apps and the browser on the smartphone. The companies manipulated so-called tracking pixels so that they could access sensitive information using the installed apps. When the study became public, browser developers like Google blocked the method, and Meta itself stated that it discontinued the trick shortly thereafter.

In short: Even though WhatsApp chats are generally encrypted, the company has its own ideas about data protection. Those who want to play it safe can switch to another service with their friends. Examples considered privacy-friendly include the US messenger service Signal, the Swiss service Threema, and the German-Swiss messenger service Wire.