Facebook scam ads with deepfakes: what an Italian court ordered Meta to do and what to save before they disappear

What the court in La Spezia ordered Meta to do

The La Spezia court ordered Meta to stop, remove and filter deepfake ads exploiting a financial adviser’s identity, in a provisional decision that a December 10 hearing will confirm or overturn.

The case: a financial adviser’s face and voice in fake ads

Scammers took a genuine video in which a financial adviser warned against scams and altered it with AI to sell fake investments. According to Corriere della Sera, the ads promised “400 euros become 12,000 in three days” and fed a pump-and-dump scheme, with damages estimated at over one million euros. A criminal complaint has also been filed, and Meta did not comment in either report.

Removal, takedown of groups and filters against reposting

Issued without first hearing Meta, the urgent decree tells Meta Platforms Ireland to stop ads, posts and pages that use the adviser’s name, image, cloned voice, biometric data or distinctive signs without consent, on Facebook, Instagram, WhatsApp and any other channel, Il Sole 24 Ore reports. Meta must remove the ads, deactivate the linked WhatsApp groups and numbers and adopt filters against reposting. The filters apply to this case only: the Digital Services Act (art. 8) rules out a general duty to monitor content.

What happens at the December 10 hearing

A precautionary decree does not decide the merits, and Meta has not been found liable. On December 10, 2026, with both sides heard, the court will decide whether the order stands and rule on the adviser’s pending requests for a daily penalty for violations and for publication of the order on Meta’s channels, according to Il Sole 24 Ore.

How deepfake scam ads on Facebook work

Deepfake scam ads on Facebook borrow a trusted face to pull people into a private chat, where the fraud takes place.

Meta scam ads run at enormous scale: the company says it removed 159 million in 2025, 92% before anyone reported them (Meta, March 2026). In the UK, a week of action led by the Financial Conduct Authority found 1,267 illegal financial ads that had reached at least 2,338,372 accounts.

From the sponsored video to the WhatsApp group

An analysis by Money.it describes the script: a sponsored video with a familiar face invites viewers into a WhatsApp group of trading “signals”, where members are told to buy the same thinly traded shares at once: a WhatsApp investment scam built on a pump and dump. Then comes a second scam, from people offering to recover the money for a fee.

How to spot a deepfake scam ad

No single sign proves an ad is a deepfake, but fake ads on Facebook often share warning signs: a famous financial figure promising fast returns, an invitation to move to WhatsApp, a sponsoring page that is new or unrelated to the person in the video, lips and voice out of sync. The complete guide to deepfakes covers more clues.

Why the ads vanish before you report them

Facebook scam ads removed as illegal lose their public record. Under article 39 of the Digital Services Act, Facebook must keep its ads in a public repository, Meta’s Ad Library, while they run and for a year afterward, but once an ad is removed as illegal or against the terms, the library stops showing its content and who paid for it. Reporting does not always help either: Il Sole 24 Ore notes that reports often go unanswered, and in a CERT Polska test 106 of 122 ads reported as fraudulent stayed up. Poland has asked the European Commission to fine Meta €250 million under the DSA (Notes from Poland); the Commission will decide. With timing this tight, TrueScreen captures and certifies the ad and its sponsoring page before the report that may wipe both from view.

What to save before the ad is removed

Save the ad, its page, its Ad Library entry, the destination and the chat, because each can disappear at a different moment.

If it is your face being used

When a deepfake ad uses your face, you need to prove that it circulated and where it ran, and to secure that proof before reporting it through Meta’s help page for impersonation or, without an account, the impostor reporting form. Fake profiles in your name are covered in the guide to Facebook impersonation evidence. For a professional whose face is being sold to strangers, TrueScreen secures the proof that the ad was out there along with a technical record of how it was captured, which an independent expert can examine.

If you have already sent money

If you have already paid, preserve the chat that persuaded you, then contact your bank and the police. The WhatsApp group has to be certified while still active, and TrueScreen certifies it through a screen recording in the TrueScreen App as you scroll. Be wary of anyone offering to recover your money for a fee: the recovery scam is the second fraud described in the guide to online trading scams and their evidence.

Mistakes that weaken the evidence

With Facebook scam ads, reporting before capturing is the classic mistake, because the report may get the ad removed. A screenshot taken outside a certified app does not show where the ad came from, when it was seen or whether it was edited; cropping it strips context, and forwarding it creates copies detached from the original. Screenshots you already have can still be certified as files: that proves their content from that moment on, but cannot show how they were taken. Victims of Facebook Marketplace scams face the same problem.

Routes to removal and redress

For Facebook scam ads, removal runs through Meta, image rights through the courts and your money through the police and your bank, and every route works better with evidence already secured.

Reporting to Meta under the Digital Services Act

Article 16 of the Digital Services Act requires platforms to offer an easy way to report illegal content, to handle the reports and to tell you their decision. Report Facebook scam ads from the three-dot menu on each ad, use Meta’s illegal content route where the law is broken, and keep a copy of your report. The DSA sets no removal deadline.

Court orders and your rights over your own image

In many jurisdictions, using someone’s image without consent can be challenged in court, and the La Spezia decree shows how fast a judge can move. Under article 9 of the DSA, judicial and administrative authorities can order platforms to act against specific illegal content. New rules on deepfakes are also under discussion in Europe, and the article on where EU deepfake law is heading follows that debate.

The police report and your bank

A police report opens the criminal side of the case, and the guide on how to report an online scam explains what to bring. Call your bank as soon as you can. Banks generally do not have to refund payments you authorized yourself the way they refund unauthorized ones; each case is different, so get legal advice.

How do you turn a scam ad that may vanish into certified evidence?

TrueScreen turns a scam ad that may vanish into certified evidence by capturing it while it is still online, with a forensic methodology that documents how and when it was acquired.

What an impersonated professional or a defrauded investor sees on screen then stays verifiable after the ad, the page or the group is gone. On a computer, the Forensic Browser, a desktop app for macOS and Windows, captures the ad as a web page along with the sponsoring page, the destination site and WhatsApp Web: each session starts with cleared data, records how the browser identifies itself to websites, checks the device clock and declares the result in the report, and keeps the page source and a web archive of each page, with network traffic capture available as an option. On a phone, the TrueScreen App certifies a screen recording, with a random code on screen tying the video to its certification. A technical report states what was captured, how, when and on which device, with the hash of each file.

For videos it did not capture at the source, TrueScreen also offers AI analysis, including deepfake detection for photos, video and audio, which you reach today in the Forensic Browser’s Media Inspector: it adds a signal, while the certified capture establishes what was online. Certification shows how and when content was captured, not that what the video says is true.

1. 1

   Capture the ad while it is still online

   Open Facebook in the Forensic Browser and capture the ad with its video, text, page name, “Sponsored” label and comments. On a phone, record it in the TrueScreen App. If the Browser Extension is all you have installed, use it now.

2. 2

   Follow the ad to the page and the destination site

   In the same session, click through to the sponsoring page and the link, so the capture shows how one leads to the next, downloads included.

3. 3

   Secure the Ad Library entry

   Search for the advertiser in Meta’s Ad Library and capture its entry before you report anything, while the library still shows the content and who paid for it (DSA art. 39).

4. 4

   Certify the WhatsApp group or chat

   Record the group in the TrueScreen App from the first message, showing its name, the number and the pitches, or open WhatsApp Web in the Forensic Browser. The guide to WhatsApp evidence in court explains how chats are used.

5. 5

   Report to Meta and hand the evidence to your lawyer

   Only now report the ad, the page and the group, and give your lawyer the certified captures and their reports. The forensic methodology captures the content at the source, verifies it and documents every step, and its last step certifies it with a qualified electronic seal (eIDAS art. 3(27)) and a qualified timestamp (eIDAS art. 3(34)), which under eIDAS carry legal presumptions of data integrity (art. 35) and of accurate date and time (art. 41), while art. 46 bars rejecting electronic evidence only because it is electronic. The court weighs the rest.

Conclusion