The wave of digital scams circulating on social media and messaging apps has forced major tech companies to take security much more seriously. It's increasingly common to receive a suspicious message, a strange link, or an unusual friend request, and a moment of inattention is all it takes for them to gain access to personal data, contacts, or even bank accounts.
In this context, Meta has decided to take action and has announced a package of new features for WhatsApp, Facebook, and Messenger designed to warn users when they are about to fall victim to a potential scam. The idea is simple but crucial: the system will raise its hand before the person clicks, accepts a request, or scans a potentially dangerous QR code.
A new layer of protection in the Meta ecosystem
The reinforcement announced by the company led by Mark Zuckerberg is based on three main pillars: advanced artificial intelligence to detect fraud patterns, alerts and preventive warnings that pop up before the interaction, and greater control over advertisers and fraudulent profiles that operate within the platforms.
Meta explains that its AI systems are now capable of analyzing text, images, and context from posts, ads, and private messages in a combined way, allowing for better identification of deceptive behavior that previously went unnoticed. One of the focuses is on celebrity and brand impersonation , a widespread tactic that attempts to leverage the image of well-known people to lend credibility to fake investments, impossible bargains, or miracle products.
At the same time, the company has adjusted its mechanisms to detect links that lead to fraudulent pages —websites that almost perfectly mimic the design of banks, stores, or official services to steal passwords or payment information. The idea is to break the chain long before the user enters any sensitive data.
This entire rollout comes after several years of social and regulatory pressure. In Europe, the Digital Services Act (DSA) requires large platforms to assess and reduce systemic risks such as online fraud, and Meta is risking not only user trust but also compliance with these regulations.
Tools to prevent scams on WhatsApp, Facebook, and Messenger
The new measures are not identical across all applications. Each of Meta's platforms receives specific features tailored to how scammers typically operate in that environment, from hijacking WhatsApp accounts to creating fake Facebook profiles.
Overall, the goal is to offer a safer digital environment and reduce the window of opportunity for cybercriminals to act. The approach is preventative: less reactive measures and more early warnings so users can take decisive action.
Another key feature is that many of these functions work automatically and silently in the background, without the user having to activate anything. However, in some cases, there is an option to grant additional permissions so that the artificial intelligence can analyze specific parts of a conversation when an alert is triggered.
By combining early detection, clear warnings, and direct options to block or report, Meta aims to reduce the number of people who even respond to a fraudulent message or follow the steps requested by the scammer.
WhatsApp: Alerts about account hijacking attempts and suspicious QR codes
In the case of WhatsApp , one of the most sensitive issues is linking accounts to new devices . In recent years, scams have proliferated in which criminals convince victims to share a verification code or scan a QR code , under the guise of contests, sweepstakes, or supposed technical problems.
To curb these practices, the app will display specific alerts when it detects attempts to link the account from a suspicious device or link . The alert will inform the user of the location from which the linking request was made and warn that it may be a scam attempt.
This way, if someone tries to hack into another person's WhatsApp account from a different phone or computer, the victim is more likely to see it coming. With a simple tap, they can block the connection or deny access , preventing third parties from entering their chats and using their identity to continue deceiving their contacts.
These measures are in addition to existing ones, such as two-step verification and regular security checks. Meta emphasizes that, while technology helps, user caution remains crucial to avoid falling into traps that begin with a message that "doesn't sound quite right."
Facebook: Alerts about friend requests and suspicious profiles
On Facebook , one of the biggest loopholes for scams continues to be friend requests from fake accounts . Starting with a simple invitation, criminals can begin asking for money, sending malicious links, or collecting private information for other fraudulent activities.
To reduce this risk, Meta is testing a system of preventative alerts when an account requesting friendship shows signs of deceptive activity . These indicators include the profile having no mutual friends, originating from a country with no apparent connection to the user, or exhibiting behaviors typical of automated accounts.
When these signals are detected, a message appears on the screen allowing you to calmly review the request and offering direct buttons to block or report the profile. The goal is to give the user more context before they instinctively click "Accept."
This type of warning can be especially useful for people less accustomed to social media or more vulnerable, such as older users, who are often targeted by identity theft scams or fake investment opportunities.
Messenger: AI that analyzes scam patterns in new conversations
The Messenger app is also getting a security boost. Meta has expanded its proactive detection tools for potentially fraudulent chats , especially those involving new contacts.
Artificial intelligence systems are trained to identify typical scam patterns in messages that include, for example, job offers that seem too good to be true, investment proposals with exorbitant returns, or urgent requests for money.
If the system detects something amiss, it will display a warning within the chat itself , explaining that the content could be part of a fraud attempt. At that point, the user can decide whether they want the AI ​​to analyze recent messages in more detail.
If the review confirms a high probability of a scam, the platform will provide information about the type of fraud and offer quick access to block or report the profile . This entire process is carried out with a balance between security and privacy, as the more in-depth analysis is only activated with the person's explicit consent.
Artificial intelligence to detect impersonation, fake links, and fraud patterns
The core of Meta's new strategy is artificial intelligence applied to security . The company claims its models can analyze millions of signals in real time to detect everything from deceptive advertising campaigns to organized networks of fake accounts.
Among the advances mentioned is the ability to locate imitations of legitimate websites that attempt to impersonate well-known banks, businesses, or services. When the systems identify that an advertisement or message is driving traffic to one of these fraudulent domains, they can take action to block it or drastically reduce its visibility.
Another key front is the fight against so-called "celebrity baiting ," a practice in which manipulated images of celebrities are used to promote dubious investments or nonexistent products. Meta has even initiated legal action against networks of fraudulent advertisers in countries like Brazil and China that used this technique on a large scale.
Thanks to a combination of automated analysis and human review teams, the company claims it can more quickly detect patterns that reveal these types of fabrications, protecting both users and the brands and public figures whose image is being used without permission.
Verification of advertisers and removal of misleading advertising
Beyond private conversations and friend requests, a significant part of the problem lies in fraudulent advertising . Ads promising impossible investments, websites selling products that never arrive, or supposed sweepstakes that are actually just a way to steal personal data.
To address this issue, Meta is expanding its identity verification system for advertisers . Companies wishing to advertise on its platforms must demonstrate who they are, what type of campaigns they run, and whether they have a history of policy violations.
The stated goal is for 90% of the company's advertising revenue to come from verified advertisers , leaving only a 10% margin for businesses considered low risk that have not yet completed the entire verification process.
This cleanup aims to reduce the space available to malicious actors who were exploiting advertising to reach large audiences with relatively small budgets. It also seeks to provide a cleaner environment for legitimate advertisers, who will no longer have to compete in auctions with campaigns of dubious origin.
Figures on fraud and regulatory pressure in Europe
The figures handled by Meta itself illustrate the magnitude of the problem. In 2025 alone, the company removed more than 159 million fraudulent ads for violating its internal policies, and closed 10,9 million accounts on Facebook and Instagram linked to cybercrime activities.
Much of this effort also stems from an increasingly demanding legal environment, particularly in the European Union . Regulations such as the Digital Services Regulation require large platforms to mitigate systemic risks , offer effective reporting mechanisms, and increase transparency regarding how advertising and potentially harmful content are managed.
In European countries, including Spain, authorities and consumer protection agencies have long warned of the proliferation of investment scams, bank impersonations, and buy-sell frauds that originate on social media or messaging apps. Meta's new measures align with this demand for greater accountability from these platforms.
At the same time, the company's collaboration with law enforcement agencies and other companies in the sector allows it to go beyond simply removing content, sharing information that can help dismantle organized networks operating on an international scale.
Taken together, this package of new features reflects a shift in Meta's approach: moving from deleting ads and accounts after the damage is done to attempting to intercept scams in real time , using contextual alerts, enhanced verification, and increasingly sophisticated artificial intelligence models. For WhatsApp, Facebook, and Messenger users in Spain and the rest of Europe, this translates into more red flags when something seems off and a slightly less comfortable ecosystem for criminals, although the usual advice remains: be wary of anything that seems too good to be true, check carefully before clicking, and use the blocking and reporting options when something seems suspicious.