Finance Watch, the Brussels-based public interest association, published a report on September 22, 2026 finding that 34 of 59 finance influencer posts reviewed across Germany, Spain, Czechia and Hungary offered only low-quality risk communication, and that the websites and apps of 24 banks, neobanks and trading platforms contained more than 100 instances of design features steering customers toward choices that may not serve them. The group wants the European Commission to use its Digital Fairness Act, due in the fourth quarter of 2026, to outlaw those interface practices and to ban influencer marketing of retail investment products outright.

In Short

A Brussels public-interest group had researchers examine 24 banking and trading apps and websites, plus 59 money-tips posts from 15 social media creators, in four EU countries. The apps kept nudging people toward particular investments with "Recommended" labels, countdown timers and price alerts, while most creator posts called themselves education and said little about how much money a viewer could lose. None of the creators was found to hold a financial advice licence, and some posts were aimed at teenagers. The group wants EU lawmakers to ban these app tricks and stop influencers promoting investment products, and the Commission's consumer law proposal due before the end of 2026 is where that argument will be settled.

A study built for a legislative window

Finance Watch commissioned The Behaviouralist, a behavioural science research company, to carry out the fieldwork between April and June 2026. According to the report, researchers ran a behavioural audit: they documented interface practices and creator content with standard templates, kept screenshots and transcripts as evidence, and classified each design practice as high, medium or low risk according to its likely effect on the consumer. Two frameworks shaped the coding. ADDICT, a qualitative risk tool published by the Arbeiterkammer Wien on February 10, 2026, was used for the interfaces, and the Behaviour Change Techniques Taxonomy for the influencer material.

Two samples were built. The first covered 24 traditional banks, neobanks and trading platforms with significant market presence in Germany, Spain, Czechia and Hungary; firms operating in several of those countries were assigned to a single national sample to avoid double counting. The report does not name the institutions, identifying them only by type and country. The second covered 15 creators, four each from Spain, Germany and Hungary and three from Czechia, chosen for audience reach, visibility and engagement. Long-form YouTube creators contributed three pieces each and short-form creators on Instagram or TikTok five each, all selected for high view counts within the preceding 12 months. That produced 59 pieces of content. According to the report's platform breakdown, 20 were TikTok videos and 15 were Instagram posts, which leaves 24 from YouTube.

The report is open about its limits. Some institutions placed features behind payment barriers or required an active trading account, and the short project window meant researchers could not capture personalised features, algorithmically tailored recommendations or notifications triggered by sustained use. Where access was restricted, the team focused on observable user journeys, promotional content and accessible interface elements. Finance Watch, whose members include consumer groups, housing associations, trade unions and NGOs, states in the report that it receives funding from the European Union to implement its work programme.

The timing is deliberate. The Commission opened a fitness check of EU consumer law on digital fairness in May 2022 and published the results in October 2024, estimating that unfair commercial practices online cost consumers at least 7.9 billion euros a year, according to the report. A consultation on the Digital Fairness Act followed in 2025. Finance Watch commissioned its study in early 2026 specifically to inform how that proposal treats financial services.

Why retail investment raises the stakes

Brussels wants more Europeans in capital markets. The Savings and Investments Union and the Retail Investment Strategy are both designed to lift retail participation, and the report sets out how much of that participation now happens on screens. In 2024, 72% of EU internet users used online banking, against 56% in 2014. In 2021, roughly 17% of EU household assets were held in financial securities, amounting to 5,610 billion euros, or 38.6% of GDP.

Information sources have moved as well. The report cites OECD research showing that in France in 2024, among new investors aged 18 to 24 and those with the lowest incomes, social media was the most commonly quoted source of information on investing at 41%, followed by social media influencers at 29%. A Commission mystery shopping exercise published in April 2022 found that 97% of the most popular websites and apps used by EU consumers deploy at least one manipulative design practice. Losses are documented too: a French consumer collective filed a complaint against about 100 influencers in January 2023 over promotions of risky investments with promises of gains, and the Dutch Authority for the Financial Markets has received dozens of complaints from consumers who lost money following influencer "tips".

Six categories of steering, counted

The report groups its interface findings into six types of dark patterns, design practices that push people toward decisions benefiting the provider. Across the 24 institutions, the instance counts were:

  • Giving more prominence to certain options: 40
  • Emotional language or shaming: 19
  • Social proofing and unverified "expert" opinions: 19
  • Pre-selected or default options: 11
  • False sense of urgency: 9
  • Hiding information about products and services: 9

Those figures total 107 recorded instances, by PPC Land's arithmetic. The report itself gives no aggregate.

The largest category combined two features. Trending lists, labelled "most traded", "most popular", "top movers", "hot" or "winners and losers", appeared in 21 instances, and risk was rated high in 14 of them because the lists were persistent, visually dominant and built into core investment screens. According to the report, these lists carry a social proof effect, exploiting the tendency to align behaviour with a larger group.

Recommendation features supplied the other 19 instances. Some drew on analysis of a customer's transaction data; others were generic suggestions shown to every user. A major German bank placed "Our recommendations" as the focal point of product pages. A neobank displayed a large "Buy" prompt beside expected return, risk-adjusted return and price trend indicators. A third interface offered "Auto Invest" and "Invest Manually" as ways to fund an investment plan, but pinned a prominent "Recommended" label on the automated, recurring option. Most of the time, the report argues, such features steer customers toward higher-fee products.

Urgency, defaults and buried caveats

False urgency surfaced nine times. A German bank's savings plan promotion, accessed on June 5, 2026, offered savings on selected ETFs and funds "from as little as" 25 euros for three years "at top conditions" until June 30, 2026, without defining those conditions. A Czech trading app ran a countdown timer on the screen where users confirmed or rejected an investment agreement.

Defaults, found 11 times and mainly in trading apps, included a pre-filled deposit of 1,000 euros in a savings plan window and quick-select buttons set at 25, 50, 100, 250 and 500 euros. None was rated low risk. Another Czech app asked users how much they planned to send each month before establishing whether they wanted a recurring plan at all, offering 2,000, 3,000 and 5,000 koruna as preset amounts.

Hidden information accounted for nine cases. A Hungarian bank put historical returns across several time horizons in the main product view, while the standard caveat that past performance does not guarantee future results sat in collapsible sections. A German bank advertised "2.7% FestzinsSparen", a 12-month fixed-rate savings product, in large green type, with an asterisk beside "p.a." pointing to conditions visible only after further clicks.

Loss framing and borrowed authority

Emotional pressure appeared 19 times, usually as loss aversion. Banners warned that inflation erodes purchasing power and presented investing as "a solution". An interactive calculator projected the purchasing power a customer would lose over five years by not investing, and another banner urged users to "build a healthy habit" by investing regularly.

Appeals to authority and crowd behaviour were also counted 19 times. One Spanish-language page carried an "analyst consensus" chart showing how 16 analysts from global investment banks and brokerages rated a stock, from "strong buy" to "strong sell". The only caveat, that the institution did not guarantee the ratings' completeness or correctness, sat in small grey text at the foot of the page and in English only. Another page offered investment ideas under the heading "What does Warren Buffett invest in?", though the options were only loosely related to Buffett and had been chosen on the basis of "emulating the expert's investment philosophy", according to the report. It cites a 2018 Commission study that found prominently displaying positive reviews or ratings almost doubles the likelihood of consumers choosing a product.

Price alerts that could not be turned off

Addictive design gets its own chapter. The most common feature was live performance updates inside mobile apps, identified 19 times: continuously refreshed prices and percentage moves coloured red or green. The report accepts that alerts can keep investors informed, but cites research suggesting notifications significantly increase trading shortly after they arrive. Risk was rated high in 15 of the 19 instances, a classification "primarily attributable to the feature's structural integration within application interfaces, which prevented users from disabling it."

On how widespread the lack of an off switch was, the text is not entirely consistent. The executive summary says that "in some cases" push notifications could not be switched off. The chapter on addictive design refers to price notifications "which could not be switched off in all cases analysed in the study", a phrase that can be read as either every case or not every case. The report also moves between "live performance updates" and "push notifications" when describing what appears to be the same feature.

Fifteen creators, 59 posts

Social media forms the other half of the study. Section 3.1 of the report says roughly 54% of the 59 pieces were presented as educational rather than as advertisements; the executive summary and conclusion describe the share as "more than 54%", and the press release as "more than half". Several creators claimed to teach financial knowledge that schools and traditional institutions do not provide.

None of the 15 finfluencers was identified as holding a formal financial advisory licence or regulated credential, and none presented themselves as a certified financial adviser. Credibility was built instead on personal investing experience, entrepreneurial success, books, past mistakes, and references to Bloomberg, Warren Buffett, JP Morgan or Jamie Dimon. That credibility-building was the most frequently coded behaviour change technique, used 28 times. Shaming and reframing, such as describing not investing as "losing money to inflation" or "working until 65 for someone else's benefit", appeared 24 times. One German TikTok creator opened a video on gold with: "The price of gold is exploding and 90% of people are just watching."

Topics clustered around ETF investing, long-term investing and stock picking. Budgeting and pension planning were the least discussed. Riskier niches, including luxury watches and precious metals, also featured, and crypto content drew particular concern: several creators discussed cryptoassets with little or no meaningful risk warning, in some cases alongside sponsorships or affiliate links to platforms such as Coinbase. A Czech Instagram creator wrote that "Bitcoin entered my TOP 3 investments. And only by growing in value."

A recurring pattern was the funnel off public platforms into Patreon, Discord, paid courses, academies and membership programmes. Content in those closed spaces is not publicly visible, which the report says places it beyond independent scrutiny by platforms or regulators. It may also amount to advertising, or even investment advice, which under MiFID II requires authorisation and an assessment of the product's suitability for the individual investor.

Risk communication by platform

Of the 59 entries coded, 34 contained low-quality risk communication, 15 medium-quality and 10 high-quality. The report defines the bands carefully. Low quality meant a generic line such as "not financial advice" or "I am not a financial advisor" without discussion of the risks tied to the content. Medium quality mentioned market risk but framed it as manageable through long holding periods or diversification. High quality put risk at the centre of the message, for instance noting that "70% of individual stock picks underperform the benchmark", or that recovering from a 90% loss could take more than 12 years even with 20% annual returns.

Format made a large difference. YouTube accounted for 9 of the 10 high-quality entries. TikTok was coded low quality in 19 of 20 videos and Instagram in 12 of 15, which by subtraction leaves only 3 of the 24 YouTube pieces in the lowest band. A Hungarian Instagram creator's disclaimer shows the typical form: "I would add that I am not a financial advisor at all; everything I know I learned from books, guides, and my own experiences." A Czech creator captioned a post "Investing is risky and past returns do not guarantee future returns. Invest sensibly", without addressing risk in the video itself. Downturns, volatility, inflation and geopolitical uncertainty were frequently recast as buying opportunities, according to the report, with little said about liquidity constraints, time horizons or the chance that markets may not recover when an investor needs the money.

Disclosure of paid relationships

The report's good-practice examples placed a label such as "Hirdetés" or "Anzeige" at the top of the caption, before any affiliate link or offer. Figure 10 reproduces a video description from the German channel Aktien mit Kopf, dated March 29, 2026, which opens with "Werbung" followed by a Coinbase offer and a risk notice.

Weaker cases fell into two groups. In 22% of the posts where a creator had a commercial relationship with the featured provider, disclosures were difficult to identify or interpret: the tie was absent from the content and caption and could only be inferred from external links, referral codes, or information in a bio or link page. In roughly 8% of such posts, no relationship was disclosed at all, despite repeated references to third-party platforms, referral links in captions or sign-up calls to action, and no #ad, #hirdetés, #publi or equivalent label. Together, those two groups add up to the 30% the report cites for posts where disclosures were hard to find.

Content aimed at minors

Several entries were directed at minors or at parents investing on a child's behalf. The report notes that TikTok's minimum age is 13 and cites an estimate that 13 to 17 year olds make up around 14% of the platform's active users globally, with 18 to 24 year olds above 30%. One Spanish TikTok video featured a 15-year-old buying an S&P 500 ETF through a custodial account: "I don't even need to be 18. My parents opened a custodial account for me." It carried no risk warning. Another Spanish TikTok creator told underage viewers who could not open an account to ask a father or guardian to open one in the adult's name, use it until turning 18, and then have the accumulated holdings transferred. The report reads that as encouraging minors to get around age restrictions and platform safeguards. A Hungarian Instagram video aimed at parents presented ETF investing for children as simple, while omitting volatility, time horizons and suitability.

Three documents, three versions of the numbers

The release came as three documents: the 38-page report, a press release, and a pitch email sent to PPC Land by Finance Watch on September 22, 2026. Their figures do not always align.

The commercial relationship statistic diverges most. The report states that "in 30% of the posts where the finfluencer had a commercial relationship with the financial services provider, disclosures were hard to find", then splits that into 22% with difficult disclosures and roughly 8% with none. The press release describes it differently: "In 30% of posts, 'finfluencers' had a commercial relationship with a financial services provider, but disclosures of these ties were often difficult to find or did not exist." The pitch email goes further, presenting 30% as the share of posts that involved a commercial relationship. In the report's framing, 30% measures poor disclosure; in the email's, it measures how much of the content was sponsored. The report does not state how many of the 59 posts carried a commercial relationship in total, so the sponsored share cannot be derived from the published text.

The recommendation feature count shifts too. The report and the press release both describe 19 instances of recommendation features across the sample. The pitch email instead says 19 of the 24 platforms used them. Instances and platforms are different units, and the report does not state how many of the 24 institutions deployed such features.

Where EU law stops short

The report tests the findings against the main horizontal and sectoral EU instruments and concludes that each covers part of the problem at most.

The Digital Services Act is the only horizontal law with a general prohibition of manipulative interface design, in Article 25. It applies only to intermediary services, however, and therefore not to banks, neobanks or trading platforms. The Commission has used the DSA against addictive design at the largest social platforms, issuing preliminary findings against TikTok on February 6, 2026 and against Instagram and Facebook on July 10, 2026, naming infinite scroll, autoplay, push notifications and recommender systems. The Finance Watch report cites the Meta findings, then points out that none of that reasoning reaches the institutions it studied.

The Unfair Commercial Practices Directive blacklists only a handful of manipulative practices in Annex I. False urgency is among them; the other practices found in the study are not, none of the listed items refers specifically to digital interfaces, and anything outside the list requires case-by-case assessment by a court. The financial services chapter of the Consumer Rights Directive, which absorbed the revised distance marketing rules, obliges member states to address only one of three listed types, and of the six types the study found, only giving more prominence to certain options falls within it. MiFID II, the Insurance Distribution Directive and the PRIIPs Regulation contain broad principles, such as acting in the client's best interests and communicating in a fair, clear and not misleading way, but were drafted before choice architecture became a policy topic and never mention it.

Influencer rules are thinner still. The Audiovisual Media Services Directive requires video-sharing platforms such as TikTok and YouTube to give uploaders a tool to flag commercial content, but does not specify when or how influencers must disclose. Its minor protection provisions, Articles 6a and 28b, were written with pornography and gratuitous violence in mind and do not identify financial promotions as harmful content. A 2021 Commission guidance notice says that influencers who endorse frequently may qualify as "traders" under the UCPD regardless of audience size. That guidance is non-binding and untested before the Court of Justice, and the report says it has produced inconsistent, "and even partly contradictory", national enforcement on labelling wording and visual prominence.

The Retail Investment Strategy, formally adopted by the co-legislators in June 2026, is the only financial services framework that addresses finfluencers by name. Investment firms remain responsible for marketing carried out on their behalf; finfluencer marketing tied to a commercial relationship must be fair, clear and not misleading; and firms that pay creators must keep records of the communications. The report lists four gaps. Independent creators without a commercial link fall outside the rules. There is no licensing or registration regime. Crypto assets and other products outside MiFID II and the IDD are excluded. And content framed as financial education, rather than promotion of specific instruments, is not treated as marketing, with no prescribed test to tell the two apart.

France has gone further on its own. Law 2023-451 of June 9, 2023 bans influencer promotion of certain risky financial products, including complex products, crypto assets and NFTs unless approved. Finance Watch regards that as a partial answer, because it leaves retail investment products with simpler risk profiles untouched.

Four recommendations for the Digital Fairness Act

All four proposals run through the UCPD. The first is a general prohibition of manipulative design, backed by an Annex I blacklist naming at least six practices: giving more prominence to certain options, false urgency, pre-selected or default options, hiding costs and risks, emotional language or shaming, and social proofing or misleading "expert" opinions. The second would oblige firms to disable by default any feature that may cause addiction-like behaviour, such as excessive push notifications, with access only on explicit opt-in. The third would add finfluencer advertising and marketing of retail investment products to Annex I, the list of practices prohibited in all circumstances. The fourth would require registration and licensing for creators offering financial literacy content, with competence standards that could draw on the EU/OECD Financial Competence Framework and record-keeping duties to support supervision.

On the ban, the report argues that disclosure labels are insufficient on their own. Even when a post is labelled as an advertisement, consumers may read an influencer's investment promotion as a personal endorsement, particularly where they like and trust the creator and have limited financial literacy.

Peter Norwood, Senior Research and Advocacy Officer at Finance Watch and the report's author, framed the argument in the press release. "Consumers seeking to invest online are confronted by a maze of dark patterns and dubious guidance on social media. The upcoming Digital Fairness Act is an opportunity to clean up, so that consumers' choices reflect their needs and appetite for risk, not the profit interests of platforms and influencers," he said. He added: "If the EU wants more retail investors, it must close the gaps in the law and give them a marketplace they can trust."

Why this matters for the marketing community

For financial services marketers, the proposals reach well beyond disclaimer copy. Several items on Finance Watch's proposed blacklist are standard conversion tools in fintech product design: ranked "most popular" lists, "Recommended" badges, pre-filled amounts and quick-select buttons, and time-limited offers. Annex I status would prohibit them in all circumstances, with no case-by-case defence. Default-off push notifications would remove one of the main re-engagement channels trading apps depend on.

Brussels has already written a version of that rule for younger users. The EU KIDS Act proposal adopted on September 17, 2026 would switch push notifications off by default for minors and lists influencer marketing among the commercial mechanics under scrutiny. The same package confirmed that a Digital Fairness Act covering manipulative practices, unfair pricing, influencer marketing and digital subscriptions is coming, alongside a revision of the Audiovisual Media Services Directive to clarify how its rules apply to influencers.

The channel at stake is growing. The report cites IOSCO research from October 2022 finding that 43% of European financial services firms planned to increase their use of influencers, the channel where those firms expected the fastest growth. An Annex I ban would close it for retail investment products across the EU, going further than the French law, and would land directly on the affiliate links and referral codes that, on this study's evidence, are where disclosure most often breaks down.

The findings sit alongside research PPC Land covered on September 11. An analysis of 1,764 English-language finance videos by Legalaes found that only 2.2% of 1,266 creators held demonstrable financial credentials and that 11.7% of videos carried a disclaimer. The methods differ sharply: Legalaes applied rule-based text classification at scale, while The Behaviouralist hand-coded 59 posts in four non-English markets. Both describe a creator population largely without formal qualifications. They part ways on YouTube, which carried the highest misleading share in the Legalaes sample yet supplied nearly all of the high-quality risk communication in the Finance Watch one.

Enforcement against creator disclosure failures has been building for months. Sweden's consumer agency documented years of non-compliance in a March 2026 report, with injunction penalties of up to 1.5 million kronor per violation. In early September 2026, YouTube said it will label brand deals that creators fail to disclose, while leaving legal responsibility with creators and brands.

Paid financial advertising already passes through a gate that organic creator content does not. Google made financial services advertiser verification mandatory in 24 more EU and EEA countries in June 2026, requiring banks, insurers and investment firms to prove their national authorisation before their ads serve. No comparable check applies to the creators in the Finance Watch sample, none of whom was identified as holding a licence. Impersonation adds a fraud dimension: on September 16, 2026, a Frankfurt court ruled that Meta could not rely on the DSA hosting exemption for fake ads using the Finanzfluss brand, a German finance education business whose co-founder's face was used to lure users into investment scams.

The advertising industry has already contested the file's scope. An IAB Europe-led coalition wrote to the Commission on July 14, 2025, arguing that the UCPD, the DSA and the GDPR already cover the ground. For financial services, the Finance Watch report argues the reverse: the DSA does not reach the providers, the UCPD blacklist does not name the practices, and sectoral rules never mention them. Whether the Commission's proposal contains a financial services blacklist, a general clause, or neither is not yet public.

What happens next

The Digital Fairness Act proposal is scheduled for the fourth quarter of 2026, according to Finance Watch. Any text would then need agreement from the European Parliament and the Council, and because Finance Watch's recommendations take the form of amendments to an existing directive, member states would have to transpose whatever survives into national law.

Timeline

Summary

Who: Finance Watch, a Brussels-based public interest association co-funded by the European Union, with the study carried out by the behavioural science consultancy The Behaviouralist and the report written by Peter Norwood. The subjects were 24 unnamed banks, neobanks and trading platforms and 15 finance influencers on YouTube, TikTok and Instagram.

What: A report recording manipulative interface design across six categories, 107 instances by PPC Land's count, 19 instances of live price updates rated as addictive design, low-quality risk communication in 34 of 59 creator posts, no identified financial licence among the 15 creators, poor or absent disclosure of commercial ties, and content aimed at minors. Finance Watch recommends a UCPD blacklist of common dark patterns, default-off addictive features, a ban on finfluencer marketing of retail investment products, and licensing for creators offering financial education.

When: Research ran from April to June 2026; the report was published on September 22, 2026, ahead of the Commission's Digital Fairness Act proposal scheduled for the fourth quarter of 2026.

Where: Germany, Spain, Czechia and Hungary, with recommendations aimed at EU-wide consumer law.

Why: The EU is pushing retail savers into capital markets through the Savings and Investments Union and the Retail Investment Strategy, while the report finds that the DSA does not apply to financial providers, the UCPD blacklist omits most of the practices observed, and the Retail Investment Strategy leaves unpaid, education-framed and crypto content outside its finfluencer rules.