Given that WhatsApp, with over three billion monthly active users worldwide, can become a fertile ground for fraud as scammers constantly refine their methods, the application owned by Meta has begun testing a new feature.
Given that WhatsApp, with over three billion monthly active users worldwide, can become a fertile ground for fraud as scammers constantly refine their methods, the application owned by Meta has begun testing a new feature.
This feature is designed to warn users about potentially fraudulent messages and help them recognize manipulation attempts. Meta stated that 'Scam Alert' is an optional feature that uses a machine learning model running directly on the device to flag suspicious messages in WhatsApp.
At the end of June, WhatsApp announced that users would no longer need to share phone numbers to communicate; instead, they could use unique WhatsApp usernames. This step was presented as a measure 'designed to protect your phone number privacy.' However, it raised concerns about fraud and brought up the possibility of scammers using usernames to impersonate people, brands, or public figures.
To alleviate concerns related to scammers, the platform is developing 'Scam Alert.' Meta emphasized that recent advancements in on-device machine learning models allow for accurate text classification entirely on mobile hardware.
Once enabled by the user, 'Scam Alert' downloads a machine learning model to the device, where it performs processes to determine if incoming messages from unregistered contacts match known scam patterns. The model is trained on patterns observed in conversations that users have sent to Meta. Classification is based on probabilistic analysis of dialogue structure and linguistic signals.
Meta assured that 'no content is automatically transmitted to WhatsApp, Meta, or any third party.' If the model identifies a message as a probable scam attempt, the user sees a warning in the chat that is not visible to the recipient. The user then decides what to do: block, report the violation, or continue the conversation.
If the user believes the warning was mistaken, they can mark the chat as trusted, after which the warning will disappear, and 'Scam Alert' will no longer flag that chat. When marking a chat as trusted, the user may also agree to share the last five received messages with WhatsApp to improve the function's accuracy.
Meta clarified that there will be no automatic reporting. The company stated: 'WhatsApp cannot initiate the exchange of any user data without user action. The only way message content or even the detection of fraud reaches our servers is if the user explicitly chooses to report it, consistent with how user reporting works in WhatsApp.' It was also confirmed that 'Scam Alert' is a user-controlled tool that can be turned on or off at any time, and all inferences and processed data remain on the device.
Currently, 'Scam Alert' is being tested in a limited beta, but Meta has not yet disclosed when the feature will be available to all users.
Meta, the owner of WhatsApp, has announced the launch of a new tool designed to protect users from cyber scammers. This tool includes an alert system.
The scam alert feature in WhatsApp is currently undergoing testing. After all trials are completed, it will be released in the stable version. Beta testing showed that the system operates on a local machine learning model and displays a red icon on suspicious messages.
The advantage of using local machine learning is the preservation of the privacy of all messages. This service functions directly on the device, and message information is not transmitted to WhatsApp, Meta, or any other platforms' servers.
Scammers have long used WhatsApp to contact victims in most cases of cyber fraud. Therefore, the new WhatsApp feature could prove to be very useful.
To activate the scam alert tool in WhatsApp, users will need to enable it manually through the WhatsApp settings menu and then automatically download the AI tool. Users retain the option to disable this feature at any time.
The alert system will run in the background of the smartphone. It will check messages received from unknown numbers. If the sender's message pattern matches scam schemes, or if the message requests personal information and demands urgent action, the feature will activate. In such a case, a warning banner will appear on the user's display.
After this warning banner appears, users can block or report the chat. Additionally, users have the right to delete the warning banner itself.
WhatsApp has begun testing in India, its main market, a feature that allows users to provide their date of birth directly within the application. This change is directly linked to the country's new data protection regulations and requirements for user age confirmation.
According to Bloomberg, Meta is investigating methods to validate age without this data being shared with third parties. This test occurs amid a growing demand for safety measures aimed at minors in digital environments.
Some Indian users have noticed a notification at the top of WhatsApp requesting the entry of their date of birth. This notice clarifies that future legislation in the country will require the platform to request the age of its users.
This initiative is tied to India's Digital Personal Data Protection Act, which establishes mechanisms to identify if an individual is under 18 when using children's data.
In response to Bloomberg, a WhatsApp spokesperson assured that the company is experimenting with alternatives with a strong focus on preserving privacy. The spokesperson stated: 'To comply with upcoming laws in India, such as the Digital Personal Data Protection Act, we are testing ways to protect privacy so that people can confirm their age.'
Currently, this new feature is in the evaluation phase and is not mandatory; WhatsApp informed that users can continue using the service even without providing their age.
Considering that India has over 600 million users, it represents WhatsApp's largest global audience, giving special attention to any modifications made in the territory, especially those related to digital security and privacy.
In recent years, governments in various nations have debated norms to restrict young people's access to social networks. Some proposals target users under 16, although Australia, for example, has not included WhatsApp in its guidelines.
Among the main obstacles faced by digital platforms is the period of intense scrutiny on Meta in India. The company had apologized after a post by Prime Minister Narendra Modi on Facebook received temporary restrictions.
Additionally, the company assured that it is improving its defenses against content related to child sexual exploitation on its platforms, including Instagram, and maintained a zero-tolerance policy for such material.
On the other hand, the Indian Ministry of Technology requested that WhatsApp postpone the launch of the username feature. This tool, which would allow people to connect without exchanging phone numbers, raised concerns about digital fraud and identity forgery.
Reinforcing the age aspect, WhatsApp emphasized that such information remains private. A spokesperson stated: 'This does not change how WhatsApp works or your experience.' The company also guaranteed that users' ages will never be shared with other individuals within the application.
