Meta on July 24, 2026 introduced Facebook Verified, a free badge that uses a recorded video selfie matched against existing profile photos to confirm that a real person sits behind a profile. Eligibility stops at accounts aged 18 and over in good standing, and the badge is not available for Pages or ProMode accounts, leaving brand and creator surfaces outside a trust signal built for Marketplace, Dating and Groups.

The announcement appeared on Meta's newsroom under the Facebook category, filed alongside Safety and Expression and tagged AI. It was published the same day as a companion post titled Connecting Real People on Facebook, and sits in a run of Facebook and Marketplace product news that also includes an app for Marketplace sellers. The framing throughout is authenticity rather than advertising.

According to Meta, the rationale is the growing difficulty of telling machine-generated identity from human identity. "As AI makes it easier to generate content, profiles, and messages, we want to provide a way for you to know there is a real person on the other side of a profile," the company stated in the announcement.

A short video, matched against old photos

The mechanism is deliberately narrow. According to Meta, Facebook Verified uses a selfie-based process in which a user records a short video selfie, which the company then checks against the profile photos already on the account to confirm a match. The stated processing time is a few minutes. There is no fee at any stage, and no government identity document is described anywhere in the announcement.

That construction is worth reading closely. The reference set is the account's own photo history rather than an external identity database, which means the check establishes continuity between a live capture and prior uploads rather than establishing a legal name. A profile with a thin or recently uploaded photo history offers a weaker reference set than one with years of images. Meta did not publish match thresholds, failure rates, appeal routes, or what happens to accounts that attempt verification and fail.

Nor did the company state what becomes of the recorded video afterwards. Retention periods, storage architecture, the presence or absence of a derived biometric template, and the legal basis for processing are all absent from the announcement text.

Eligibility narrows the field twice

Two exclusions do most of the structural work.

The first is age. According to Meta, Facebook Verified is available to eligible Facebook users 18 years and older who are in good standing with Community Standards, specifically those prohibiting fraud, scams and deceptive practices, and who exhibit no evidence of inauthentic behavior. Accounts must meet the company's trust and safety standards to qualify. A user already flagged for deceptive activity is therefore outside the programme by design, which limits the badge's usefulness as a fraud filter to the population that was never flagged in the first place.

The second exclusion carries the commercial weight. According to Meta, the badge is not available for Pages or ProMode accounts. Pages are the surface brands, publishers and businesses use on Facebook. ProMode is the professional configuration for individual creators. Both sit outside the free badge entirely, which places Facebook Verified squarely on the personal-profile layer and leaves commercial accounts where they were.

Rollout is phased. According to Meta, the company is starting in select markets with plans to expand globally. No market list accompanied the announcement, and no timetable was attached to the global expansion.

Marketplace, Dating, Groups, Profile, then Feed

Placement follows the same logic. According to Meta, once verified, the badge appears across the places on Facebook where it matters most: Marketplace, Dating, Groups and Profile to start, with badges added in Feed posts over time. Verification happens once, and the badge travels with the account across the service.

The ordering is telling. Marketplace, Dating and Groups are the three surfaces where a stranger's identity carries transactional or physical consequence, and all three are surfaces where fraudulent behaviour has been documented at scale. Feed, the surface that carries the advertising inventory, comes later.

Meta describes the intended moment of use as pre-commitment: the badge is meant to be visible while browsing Marketplace listings, looking at a dating profile, or checking a profile inside a Group.

What the badge is not

The announcement contains an unusually explicit disclaimer. According to Meta, the Facebook Verified badge does not mean that Facebook endorses the user or guarantees their trustworthiness, and all users remain subject to Meta's policies including Community Standards and Commerce Policies.

That distinction separates two claims that consumers routinely collapse. A liveness-checked face match can establish that a human completed a step. It cannot establish that the human intends to ship the item, appear at the meeting, or represent the goods accurately. A verified seller who takes payment and disappears has not defeated the verification system, because the system never made a claim about conduct.

A free tier arrives beside a paid one

Meta already sells verification. Meta Verified launched as a paid subscription in 2023, providing a blue badge, impersonation protection and account support, and expanded to businesses on Facebook and Instagram in India in July 2024 with four tiers running from roughly 639 rupees per month to 30,000 rupees for the most comprehensive plan. Premium platform capabilities have been attached to those tiers since, with clickable links in Instagram post captions surfacing as a tier-gated test in March 2026.

Facebook Verified does not replace that product. It runs parallel to it, at zero cost, on a different account class, answering a different question. The paid badge answers whether an account is the authentic presence of a named business or public identity. The free badge answers whether a human completed a face check. Businesses that want a visible trust marker are still routed toward the subscription.

The pattern is not confined to Meta. Reddit began alpha testing verified profiles with grey checkmarks in December 2025, restricting selection to public figures, experts and trusted partners rather than opening requests to all users. Identity marking has become a standard platform layer rather than a differentiator.

The fraud record the badge sits on top of

Context matters here, because Facebook Verified arrives into an enforcement environment that has been unusually hostile to Meta's authenticity claims.

Internal company documents published by Reuters in November 2025 and covered in detail at the time estimated that Meta platforms exposed users to roughly 15 billion higher-risk scam advertisements each day, alongside 22 billion organic scam attempts daily as of a December 2024 internal presentation, a figure that explicitly covered fraudulent Marketplace listings, fake dating profiles and phoney medical treatments promoted in health groups. Those are the exact three surfaces Facebook Verified now badges first.

The same materials recorded that Facebook and Instagram users filed approximately 100,000 valid reports each week about fraudsters messaging them, and that 96% of those reports were ignored or incorrectly rejected. Enforcement pressure followed. IAB Sweden expelled Meta from its membership in March 2026 over insufficient action on deceptive advertising. The Consumer Federation of America filed a class action in the Superior Court of the District of Columbia on April 21, 2026, arguing that the auction system financially rewarded lax enforcement.

Meta's own countermeasures have leaned on the same technology now powering the badge. In an announcement published on March 11, 2026, the company reported removing over 159 million scam ads during 2025, with 92% taken down before any user report, and set a target of routing 90% of advertising revenue through verified advertisers by the end of 2026, up from roughly 70%. Facial recognition deployed against celebrity impersonation had more than doubled detection volumes during testing.

Read against that record, Facebook Verified extends a verification logic already applied to advertisers down into the consumer profile layer, on the peer-to-peer surfaces where Meta's own documents located the largest volume of organic fraud.

Biometrics arrive as the rulebook tightens

Face-based verification is now a contested regulatory category, and the timing is conspicuous. Google introduced selfie video sign-in for Google Accounts on July 23, 2026, one day before Meta's announcement, with Help Center documentation disclosing regional limits, an undefined retention period, longer retention for policy violators, and an optional setting allowing the footage to train facial recognition and age estimation systems.

Biometric data falls under Article 9 of the GDPR, which imposes a general prohibition on processing absent explicit consent or another named legal basis, with penalties reaching 20 million euros or 4% of global annual turnover. Recent decisions show what regulators penalise. Spain's AEPD fined age assurance vendor Yoti 950,000 euros in March 2026, including 500,000 euros for unlawful processing of biometric special category data and 200,000 euros for invalid consent obtained through pre-ticked checkboxes. The same authority fined airport operator AENA 1.8 million euros in November 2025 for inadequate impact assessments before deploying facial recognition, establishing that valid consent does not exempt controllers from assessment duties. The Article 9 framework and its penalty ceiling were set out in detail in coverage of the PimEyes case.

Brazil's ANPD opened a consultation in May 2026 on a draft age verification guide that separates verification, estimation and recognition, describing recognition as the conversion of a face image into a template compared against a stored reference. A selfie matched against stored profile photos falls into that category, which carries the heaviest legal load.

Meta is already inside European proceedings on adjacent grounds. The European Commission adopted preliminary findings on April 29, 2026 concerning age assurance for users under 13, and issued further preliminary findings on July 10, 2026 on the addictive design of Instagram and Facebook, naming infinite scroll, autoplay, push notifications and personalised recommender systems, with exposure of up to 6% of global turnover under the DSA. Meta had already extended AI-based age detection to the EU, Brazil and US Facebook on May 5, 2026, using visual and contextual signals. The absence of a named market list in the Facebook Verified announcement is consistent with a deployment sequenced around those constraints.

Why this matters for marketers

Three consequences follow directly from the design.

The exclusion of Pages is the first. Brands, publishers and businesses operating on Facebook cannot obtain this badge, so any authenticity gap between a badged private seller and an unbadged commercial Page on Marketplace runs against the commercial account. Where peer sellers and brand storefronts compete inside the same listing surface, the visual trust marker is available only to one side.

The second is the shifting baseline for organic reach on peer surfaces. Marketplace has been accumulating commercial infrastructure, with collections, collaborative buying and AI-generated listing insights added in November 2025. A verification badge layered onto that infrastructure creates a soft ranking signal for buyer attention that is not purchasable and not open to Pages.

The third is the wider authenticity economy. Detection vendors have already commercialised the problem on the open web, with IAS making AI slop avoidance generally available in June 2026 and Pangram data showing over 40% of top-level LinkedIn posts fully AI-generated. Research has also found that roughly half of users cannot reliably distinguish AI-generated accounts from human ones. Platform-issued human badges and third-party synthetic-content detection are converging on the same problem from opposite ends, and Feed placement, when it arrives, will put that signal next to advertising inventory for the first time.

What remains undisclosed is substantial: no market list, no retention policy, no legal basis, no controller entity, no accuracy figures, and no date for Feed badges. According to Meta, the purpose is to keep the moments that matter on Facebook grounded in real people. The company has published what the badge asserts. It has not yet published what the system stores.

Timeline

Summary

Who: Meta Platforms, through a post published on the company's newsroom under the Facebook category, affecting eligible personal Facebook accounts aged 18 and over. Pages and ProMode accounts, the surfaces used by businesses, publishers and professional creators, are excluded.

What: Facebook Verified, a free badge that confirms a real person is behind a profile through a recorded video selfie matched against the account's existing profile photos. The process carries no subscription fee, takes a few minutes according to the company, and requires accounts to be in good standing with Community Standards prohibiting fraud, scams and deceptive practices, with no evidence of inauthentic behavior. Meta states the badge does not signify endorsement or guarantee trustworthiness, and that all users remain subject to Community Standards and Commerce Policies.

When: July 24, 2026, with a phased rollout beginning in unnamed select markets and no published timetable for global expansion or for the later addition of badges in Feed posts.

Where: Facebook, with the badge appearing initially on Marketplace, Dating, Groups and Profile, and in Feed posts at an unspecified later date.

Why: Meta attributes the launch to the growing ease of generating content, profiles and messages with artificial intelligence, stating that a clear signal distinguishing real people has become essential on the surfaces where users transact, meet and converse with strangers.