Meta Expands AI-Driven Anti-Scam Protections Across Facebook, Messenger, and WhatsApp
Meta announced expanded anti-scam measures across Facebook, Messenger, and WhatsApp, emphasizing AI-driven detection of impersonation, deceptive links, and other fraud patterns, alongside new user-facing warnings intended to interrupt scams earlier in the interaction. Updates include Facebook alerts for suspicious friend requests, WhatsApp warnings for potentially fraudulent device-linking attempts (e.g., QR-code based linking), and Messenger prompts that can offer an AI scam review of recent chat content; Meta also said it is expanding advertiser verification to reduce identity misrepresentation in ads. Separately, Meta described enforcement at scale, reporting the removal of 159 million scam ads in 2025 and 10.9 million Facebook/Instagram accounts tied to criminal scam centers, amid ongoing scrutiny from US lawmakers and reporting that has questioned the platform’s financial incentives to police fraudulent advertising.
Meta also highlighted collaboration with law enforcement targeting “industrialized” scam operations, including actions tied to Southeast Asian scam compounds that resulted in 21 arrests and the disabling of 150,000+ accounts, as well as broader efforts to counter “pig-butchering”-style investment fraud. Complementing these initiatives, Meta detailed a privacy-preserving Messenger capability—Advanced Browsing Protection (ABP)—that warns users about potentially malicious websites opened from encrypted chats by using cryptographic/private-information-retrieval techniques to check links against large blocklists without revealing message contents to Meta. In parallel reporting on the scam ecosystem, researchers described a large network of paid Meta ad campaigns using fake media brands and impersonated public figures to push investment scams across dozens of countries, underscoring the continued role of malvertising and disinformation-for-profit tactics in driving victim acquisition.
Timeline
Mar 20, 2026
Meta says AI support and moderation systems are rolling out globally
Meta announced a global rollout of AI systems for customer support, moderation, and security tasks, including scam detection and account-takeover identification. The company said early tests showed one tool mitigating about 5,000 daily password-stealing scam attempts and another reducing reports of fake celebrity profiles by more than 80%.
Mar 11, 2026
UK announces Online Crime Centre to fight scam-compound fraud
The U.K. government announced a new Online Crime Centre backed by more than £30 million to combat cybercrime and scam-compound-driven fraud. The initiative plans to use AI and scam-baiting chatbots while coordinating with law enforcement, intelligence agencies, financial institutions, telecoms, and major tech firms.
Mar 11, 2026
Meta expands advertiser verification target through 2026
Meta said it is expanding advertiser verification and set a goal for 90% of ad revenue to come from verified advertisers by the end of 2026, up from about 70% currently. The measure was presented as part of its broader anti-scam and ad-integrity strategy.
Mar 11, 2026
Meta announces new AI anti-scam protections across its apps
Meta announced new anti-scam tools for Facebook, Messenger, and WhatsApp, including suspicious friend-request alerts, warnings during WhatsApp device-linking flows, and expanded Messenger scam detection with optional AI review of recent chats. The company said its AI analyzes text, images, and contextual signals to detect impersonation scams, deceptive links, and domain spoofing.
Mar 11, 2026
Thai-led operation arrests 21 and Meta disables 150,000 scam accounts
Meta said a recent joint operation with the Royal Thai Police and international partners including the FBI, UK National Crime Agency, and Australian Federal Police led to 21 arrests. As part of the same disruption, Meta disabled more than 150,000 accounts linked to Southeast Asian scam-center networks.
Mar 10, 2026
Meta launches private link-checking protection in Messenger
Meta described an "Advanced Browsing Protection" feature for Messenger that warns users before opening potentially harmful links shared in chats. The system uses privacy-preserving cryptography to check links against an expanded malicious-site watchlist without revealing readable URLs to Meta despite end-to-end encryption constraints.
Feb 1, 2026
Bitdefender identifies large Meta ad investment-scam operation
Bitdefender researchers reported that a coordinated malvertising network had been running since at least February 2026, using more than 300 paid Meta ad campaigns to promote investment fraud in at least 25 countries. The operation allegedly used impersonated news brands, public figures, redirects, and lookalike domains to funnel victims into call-center-driven scams.
Dec 31, 2025
Meta discloses 2025 scam enforcement totals
Meta reported that during 2025 it removed 159 million scam ads and took down 10.9 million Facebook and Instagram accounts linked to criminal scam centers. The company framed the figures as evidence of increasingly industrialized online fraud and a shift toward disrupting entire scam networks.
Dec 1, 2025
Meta pilot disruption removes 59,000 scam-linked assets
In a December 2025 pilot operation, Meta removed 59,000 accounts, Pages, and Groups tied to scam activity. The action also contributed to six arrest warrants, serving as an earlier phase of the company's broader crackdown on Southeast Asian scam centers.
See the full picture in Mallory
Mallory subscribers get deeper analysis on every story, including:
Who’s affected and how
Deep-dive technical analysis
Actionable next steps for your team
IPs, domains, hashes, and more
Ask questions and take action on every story
Filter by topic, classification, timeframe
Get matching stories delivered automatically
Related Entities
Organizations
Affected Products
Sources
5 more from sources like zdnet zero day, register security, wired com security, the record media and help net security
Related Stories
Meta Expands Anti-Scam Protections Across WhatsApp, Facebook, and Messenger
**Meta** introduced new anti-scam protections across *WhatsApp*, *Facebook*, and *Messenger* to counter fraud campaigns that rely on social engineering, impersonation, and malicious links. The updates include WhatsApp warnings when device-linking requests show scam-related behavioral signals, such as attempts to trick users into sharing linking codes or QR codes, and Facebook alerts for suspicious friend requests from accounts with indicators like recent creation or no mutual connections. Messenger is also adding AI-driven scam detection to identify patterns associated with impersonation and spoofed links in chats. The changes are part of a broader anti-fraud push in which Meta said it worked with international law enforcement to disable more than **150,000 scam-linked accounts** and support the arrest of **21 individuals**. A separate report on a new cross-industry anti-scam accord involving Meta, Google, Microsoft, Amazon, OpenAI, and others describes a wider effort to share threat intelligence, improve fraud reporting, strengthen transaction verification, and coordinate defenses against scam operations that move across multiple online platforms. A report on **Operation Atlantic** focuses instead on cryptocurrency approval-phishing enforcement by U.S., U.K., and Canadian authorities and is a different story from Meta's platform-specific product rollout.
1 months ago
Meta Expands Safety and Enforcement Measures Across Facebook and Instagram
Meta disclosed a set of new **platform safety and enforcement actions** aimed at reducing harm and abuse on its services. The company filed multiple lawsuits against alleged scam-ad operators in **Brazil, China, Vietnam** and elsewhere, describing tactics including **deepfakes/celebrity impersonation**, “celeb-bait” investment lures, and **cloaking** used to evade ad review; Meta said it also took technical steps such as disabling accounts, suspending scam-linked payment methods, and blocking associated domains, and shared information with industry partners to help them block the same actors. Separately, Meta announced new **Instagram parental-supervision alerts** that notify parents when a teen repeatedly searches for **self-harm or suicide-related terms** within a short time window (initially for supervised accounts in the **U.S., U.K., Australia, and Canada**), and said it is developing similar notifications for teens’ **AI-related conversations** about self-harm. In parallel regulatory developments, EU lawmakers advanced a non-binding opinion supporting **privacy-friendly age verification** and proposing restrictions that would require **parental consent for under-16s** and bar access for children under 13, positioning these measures for potential inclusion in a future **Digital Fairness Act** focused on child protection online, targeted advertising, and addictive design patterns.
1 months ago
Social Platforms Face Scrutiny Over Scam Advertising and Messaging-Facilitated APP Fraud
**Meta** said it is pursuing legal action against advertisers it alleges ran **celeb-bait scam ads** on its platforms, including defendants based in **Brazil, China, and Vietnam**. Meta reported suspending the advertisers’ payment methods, disabling related accounts, and blocking domains used in the campaigns, which allegedly misused celebrity images/voices (including synthetic media) to drive victims to scam sites that harvest sensitive data or solicit money via fraudulent products and investment schemes; Meta also said it sent cease-and-desist letters to marketing consultants promoting services to bypass ad-policy enforcement (e.g., “un-ban” services and renting access to trusted accounts). Separately, a **Revolut** report on **Authorised Push Payment (APP) fraud** found scam origination increasingly tied to encrypted/direct messaging channels, with **Telegram** exceeding **20%** of reported scam origination and showing significant growth versus 2024, while **Meta-owned platforms** collectively remained the largest source at **44%**. The report highlighted category concentration by platform, including Telegram and WhatsApp accounting for major shares of **investment** and **job scams** (with Telegram cited as over **58%** of job scams), reinforcing that a large portion of financial fraud is initiated on social and messaging platforms and prompting calls for stronger platform accountability.
1 months ago