Visa's fake AI lamb posts unmasked themselves to teach scrollers about fraud

Visa wanted to raise awareness of the AI techniques used by online fraudsters and restore trust in online commerce, as part of its Europe-wide initiative 'Can You Trust What You See Online'.

Use fraudsters' own psychological triggers in reverse: deliberately post AI fakes, then reply to everyone who takes the bait with a personalised reveal.

Paris agency Marcel posted two weeks of outlandish AI-generated images and videos from the anonymous account Petit Pamplemousse — including the plight of a rare, feathered breed of lamb — passing them off as real. Once users commented, a reply revealed the trick with a personalised video and directed them to a Visa website of practical advice on AI and online fraud plus Visa's own payment-security solutions.

In its first few days the campaign recorded over 500,000 interactions, with the AI posts drawing more than 20,000 comments, and reached over 2 million people.

Visa's 'The Feathered Lamb' campaign starts from an uncomfortable fact: the vast majority of scams on social media originate from fake ads, which AI tools are making increasingly hard to detect. Rather than warn people about it, Paris-based agency Marcel decided to run the scam itself — with the victim's consent retroactively built in.

Using the anonymous account Petit Pamplemousse, Marcel deliberately deployed a series of AI-generated images and videos passed off as real, introducing outlandish news stories such as the plight of a rare, feathered breed of lamb. The posts sat on social media for two weeks while users reacted. Once someone took the bait and left a comment, the trick was revealed: a reply carrying a personalised video exposing the fakery and directing them to a Visa website with practical advice on AI and online fraud, plus details of Visa's own solutions for mitigating fraud during payment journeys.

'We wanted to use the same psychological triggers as fraudsters but in reverse,' said Julian Clobus, Visa's brand and product marketing director. 'Transaction security is at the core of our mission, but the main vulnerability remains human.' In its first few days the campaign recorded over 500,000 interactions and more than 20,000 comments, reaching over 2 million people — attention earned, as Creative Review put it, by flipping fraudsters' tactics on them.

Awareness campaigns about scams are usually ignored; this one made each participant the protagonist of their own near-miss.

The reveal is personalised and arrives at the exact moment the lesson lands — right after the person has believed a fake.

Demonstrating how convincing AI fakes have become says more about the threat than any statistic could.

The bait-to-education pipeline funnels engaged users straight to Visa's fraud advice and product solutions.

To teach people how deception feels, let them be deceived once — safely. A personalised reveal turns the embarrassment of falling for it into attention, which a warning never gets.

The campaign runs as part of Visa's Europe-wide initiative 'Can You Trust What You See Online', aimed at raising awareness of the new AI techniques used by fraudsters and restoring trust in online commerce.