Common online marketplace scams and how to recognise them
How the most common scams against marketplace buyers and sellers work, the red flags that give them away, and where to report them in the UK, EU and US.
How an Amsterdam court ordered Facebook to block fake crypto ads using a celebrity's face in 2019, and where platform duties and consumer defences stand now.
From late 2018, Dutch users of Facebook and Instagram kept seeing the same kind of advert: a famous face, a breathless story about a secret Bitcoin fortune, and a link to an “investment platform”. One of the people whose name and photo were being misused, the media entrepreneur John de Mol, took Facebook to court. The ruling he obtained in November 2019 was one of the first in Europe to order a platform to take preventive steps against a specific pattern of scam ads, and the problem it addressed is still with us.
According to the facts set out in the judgment, adverts associating well-known Dutch people, including De Mol, with Bitcoin investments appeared on Facebook and Instagram from October 2018 until at least the second half of March 2019. De Mol’s photo and name were used without his consent. People who responded and paid money received no Bitcoin and lost what they had paid. The court recorded that, at the time the case was brought, the Dutch Fraud Helpdesk (Fraudehelpdesk) had received reports of €1.7 million in losses linked to these ads.
De Mol had publicly distanced himself from the ads in a statement on 31 October 2018, and his staff reported them to Facebook. He issued a writ on 30 April 2019, and the case was heard on 5 June 2019.
The case (ECLI:NL:RBAMS:2019:8415) was decided in summary proceedings (kort geding) by the preliminary relief judge of the Amsterdam District Court, against Facebook Ireland Limited and Facebook Netherlands B.V. The court ordered Facebook, within five days of service, to stop allowing adverts on Facebook and Instagram in which De Mol’s name or image, either in the ad or on the website it led to, was linked with Bitcoin or other cryptocurrencies, and to keep them off. Non-compliance carried a penalty of €10,000 per day, capped at €1 million.
Facebook Ireland was also ordered to hand over, within seven days, the usage and payment data it held for the advertiser payment accounts identified in the evidence, with a separate penalty of €1,000 per day up to €100,000. Facebook was ordered to pay costs. The court added that no penalties would be incurred if it would be unreasonable to expect more effort and care from Facebook than it had shown. The court’s press release summarised the outcome for the public.
Three parts of the judgment explain why it attracted attention well beyond the Netherlands.
No hosting shield for ads. Dutch law, implementing the EU e-Commerce Directive, protects intermediaries from liability for content they host if they lack knowledge of it or act promptly once aware. The court accepted that Facebook was in principle an information society service but held, provisionally, that it could not rely on the exemption here. Advertising was Facebook’s primary revenue model; it set prices and actively decided which ads appeared. In this context it was acting as an operator of advertising space, not as a neutral communication platform. The court added that even where the exemption applies, it does not prevent injunctions, including forward-looking ones.
A targeted filter is not general monitoring. Facebook argued that the order would amount to a general monitoring obligation, which the directive prohibits. The court disagreed because De Mol had narrowed his claim to ads linking his name or image with cryptocurrency. It cited the Court of Justice’s judgment of 3 October 2019 in Glawischnig-Piesczek v Facebook Ireland (C-18/18), which confirmed that national courts may order hosts to remove or block specific unlawful information.
Imperfection is not a defence. Facebook said its screening could never catch everything. The court accepted that this might be true but held that an order does not require proof in advance of 100% effectiveness, nor must it dictate exactly which technical measures to use.
Facebook announced on 6 December 2019 that it would appeal, citing technical and legal difficulties in carrying out the order, NOS reported. We have not relied on any account of the appeal’s outcome, which we were unable to confirm from a primary source.
The “celebrity bait” format, a trusted face attached to a fake investment story, outlived the 2019 case. When Meta announced new countermeasures five years later, it described scammers using images of public figures to lure people to scam websites as a continuing problem across its services. In our assessment, the ruling showed that a determined individual could force targeted action for one name, but it could not protect the many public figures, and far more consumers, who lacked the means to litigate.
On 21 October 2024 Meta announced tests of facial recognition against celeb-bait ads. Where its systems suspect an ad is a scam featuring a public figure, Meta compares the faces in the ad with that person’s Facebook and Instagram profile pictures and blocks confirmed scams. Meta says it deletes facial data from the ad immediately after this one-time comparison. An update on 4 March 2025 extended the measures to the UK, the EU and South Korea, with public figures in the UK and EU able to opt in. In a further update in September 2025, Meta said user reports of celeb-bait ads, relative to total ad impressions, fell by 22% globally in the first half of 2025, and that nearly 500,000 public figures were covered. These are Meta’s own figures and have not been independently verified.
The Digital Services Act moved much of what De Mol had to litigate for into general law. All online platforms must run notice-and-action systems and show users who is behind an advert. Very large online platforms, which include Facebook and Instagram, must also assess and mitigate systemic risks arising from their services, including their advertising systems, and keep public repositories of the ads they run. On 30 April 2024 the Commission opened formal proceedings against Facebook and Instagram covering, among other things, Meta’s policies and practices on deceptive advertising. Opening proceedings does not prejudge the outcome; readers should check the Commission’s DSA pages for the current status.
Section 38 of the Online Safety Act 2023 requires providers of the largest user-to-user services, designated Category 1, to use proportionate systems to prevent people encountering fraudulent paid-for adverts, to minimise how long such ads stay up, and to take them down swiftly once alerted. Section 39 sets similar duties for the largest search services. The section is formally in force, but in practice the duty depends on Ofcom designating services and issuing codes of practice. Ofcom published its register of categorised services and consulted on draft fraudulent advertising codes in July 2026; the consultation closed in October 2026. As of November 2026 final codes had not been issued, and they were not expected before 2027.
The basic tactics have changed little since 2019, although AI-generated video and audio can now make a fake endorsement look and sound more convincing. Warning signs include:
If you see one, report it through the platform’s own ad reporting tool. In the UK you can also use the ASA’s scam ad reporting form; the ASA says it passes reports to platforms and ad networks and shares them with the National Cyber Security Centre. Before investing with any firm in the UK, check it against the FCA’s ScamSmart resources and Warning List. If you have already paid, contact your bank immediately, then report to the police or national fraud reporting service; in the Netherlands, the Fraudehelpdesk takes reports.
Businesses are targets too: impersonation ads can use a retailer’s brand as easily as a celebrity’s face. Our guides to common online marketplace scams and building customer trust as a small online retailer cover how to warn customers, and our analysis of EU digital-platform rules for online sellers explains the DSA’s notice-and-action routes.
The 2019 judgment established, at least in Dutch summary proceedings, that an ad-funded platform can be told to filter a narrowly defined category of scam ads, and that imperfect technology is no answer to such an order. Seven years later, legislation in the EU and UK has turned parts of that logic into standing duties for the largest platforms, but the scams persist. For consumers, the safest assumption remains the simplest: a celebrity in an advert promising easy investment returns is almost certainly not endorsing anything.
This article is general information, not legal advice.
How the most common scams against marketplace buyers and sellers work, the red flags that give them away, and where to report them in the UK, EU and US.
Europol's December 2019 IOS X operation against counterfeit and piracy domains, what research shows since, and how shoppers and retailers can respond.
Data minimisation, access control, patching, backups, admin MFA, logging, vendor risk and breach response for small online shops, with a 30-day checklist.