Meta research

Anonymous third-party paid political advertising in the Cyprus 2026 parliamentary elections

MedDMO researchers tracked ads naming Cyprus’s 753 registered candidates across Meta, Google and TikTok ad libraries/APIs and identified five anonymous third-party advertising operations totalling more than 450 political ads, including one spreading fabricated polling data days before the vote. Enforcement was weak and reactive: Meta removed less than 1% of flagged ads while still live and TikTok logged only one removal in 220 days, but after researchers escalated directly to the platforms and the European Commission, Meta removed 687 ads within 30 hours and TikTok removed 631 ads within five days.

Ex-post analysis of the rapid response system during Malta’s 2026 general elections

Over a 21-day window around Malta’s snap 30 May 2026 general election, MedDMO’s Maltese team reported 296 URLs to Meta and Google, the majority unlabelled political advertising in breach of EU ad-transparency rules and the platforms’ own terms. Platforms mostly acted within one to two days, but enforcement was inconsistent: on Google, removed unlabelled ads were sometimes almost immediately republished by the same advertisers, and on Meta a coordinated inauthentic network of pages/profiles was taken down, partially reinstated on election day, then removed again.

Online Domestic Extremism in Canada Data Briefing (Dec 2025-Feb 2026)

Recurring data briefing tracking violent-extremist activity across major platforms in Canada; consistently finds Telegram the leading platform for such activity.

Safety Off: threats and abuse surged after Meta's policy changes

Analysis of ~8M Facebook comments on Members of Congress found violent threats quadrupled after Meta's Jan 2025 rollback.

Analysis of the Fraud Ecosystem of Online Advertising on Meta Platforms

The ÖIAT analyzed Meta's Ad Library over three months and found 634,000+ fraudulent or problematic ads shown more than 1 billion times across the EU (123M in Austria). Illegal gambling (~450,000 ads), investment fraud (83,216) and dubious food supplements (27,171) dominated the sample.

Pornographic Advertising on Meta Platforms Reaches Millions of Users Across Europe

Online Risk Labs has identified 25 large clusters of advertising accounts that systematically run advertisements promoting pornography and other adult content across the European Union and Great Britain.

Impact of regulation of children’s digital lives – Phase II

Researchers recorded 108 child-safety/privacy design changes across 70 platforms between 2024 and 2026, but only 5 of the platforms contacted directly responded to information requests. The four largest platforms (Meta, Google, TikTok, Snapchat) have shifted away from protective ‘by default’ design changes toward optional end-user tools such as parental controls, reversing the trend identified in the Phase I report.

Scambook: How Meta helps Medicare scammers target seniors

Meta earned ~$14.3M from deceptive scam ads (including deepfakes) targeting seniors, violating its policies.

The Infinite Slop Machine

Documented a coordinated influence operation of at least 29 YouTube accounts publishing over 7,300 AI-generated geopolitical videos (March-December 2025), producing roughly two videos a day with AI narration and synthetic thumbnails to push pro-BRICS, anti-EU and anti-Western narratives.

Online Fraud in Czechia: No Improvement in Sight

Online fraud remains a multi-billion-crown problem in Czechia, and the problem continues to grow. Fraudsters systematically exploit the lack of transparency in the online advertising market, while major platforms continue to lag behind in addressing this issue.

Clickbait Cures: Unmasking the Deceptive Health Ad Infrastructure on Meta and Google

Collected 350,549 Facebook ads and 2,073 Google ads (2023-2026) promoting unregulated nutraceuticals with false medical claims, reaching an estimated 878 million accounts across the EU on Facebook; 20% of the 390 products studied had been flagged by EU or national health authorities as illegal or dangerous, yet coordinated advertiser networks kept promoting them.

Online Hate & Harassment Targeting Public Figures and Influencers on Instagram in Jordan

Analyzes the scale and nature of hate and harassment directed at public figures and influencers on Instagram in Jordan.

Information Laundering in Slovakia: An Analysis of Russian Influence Operation Networks

Mapped roughly 253 disinformation outlets and over 1,800 pro-Kremlin Facebook pages in Slovakia that translate and relaunder sanctioned Russian content through local-language intermediaries, and identifies Slovakia as the EU’s weakest enforcer of sanctions against these operations.

Weaponising social media: smears of Indigenous leaders in Guatemala

Deregulation of major platforms is helping drive online smear campaigns and criminalisation of Guatemalan land and democracy defenders.

Breaking the Loop: Exploring User Agency in Escaping Algorithmic Rabbit Holes on Social Media

Early-stage findings from a Panoptykon/Piotr Sapieżyński (Northeastern University) study testing whether Facebook's 'Not interested' feedback tool lets users actually escape unwanted recommended content; results are mixed, with one pilot showing no lasting effect and a later 7-participant study showing a 38% average reduction with wide variance.

Uneven Moderation by Language: Differences in DSA Content Removals in 2025

Companion study finding that content reported in smaller or minority EU languages was removed far less often (23.3%) than content reported in major languages (40.2%), pointing to language-based gaps in how platforms moderate under the DSA.

Uneven Moderation Across Platforms: Differences in DSA Content Removals in 2025

Of 13,187 posts and comments reported under the DSA across seven EU languages in 2025, 34.6% were removed overall, but action rates ranged from 71.3% on YouTube down to 12.5% on X, a roughly 5.7x disparity that evidences sharply uneven DSA enforcement between platforms.

'Your report is important to us': unnecessary barriers to submit content reports

Analyzes how platform content-reporting mechanisms impose unnecessary barriers under the DSA.

Brief for the Hearing at the European Parliament on DSA Enforcement and the Protection of Minors

Prepared for an EPP/Renew-organised European Parliament hearing, this brief matches TikTok, Meta and YouTube's self-declared minor-safety mitigation measures against independent research and finds none of the platforms disclose data substantiating that their measures actually work.

Sexualized images on X: what we are doing to stop them and what we expect from the EU

Documented networks of X accounts openly promoting non-consensual 'nudify' tools; a reported post was deemed non-violating, and X refused a DSA Art. 40.12 data request, blocking research (context to the EU's DSA fine on X).

War profiteers: online ads and the machinery of propaganda for war

Examines how governments deploy war propaganda through online advertising and argues platforms bear responsibility under international human rights law to prevent monetizing it.

Malicious semi-compliance as the platform approach to EU regulation

Argues big tech feigns compliance with the DSA while sidestepping its intent through 'malicious semi-compliance.'

Hate for Sale: Instagram helps sellers of hateful merchandise

Instagram and e-commerce platforms amplify and monetize racist/antisemitic 'hate merch,' reaching a billion views.

Still marching on(line): How R-FBI Targets Moldova’s Elections

Detailed how the Russian operation Storm-1516 (linked to the sanctioned R-FBI) used fabricated news sites, AI-generated content and impersonated journalists to push 17 false narratives against President Sandu and the PAS party during Moldova’s 2025 elections, accumulating roughly 45 million views via authentic influencers and inauthentic accounts.

A Cross-Country Analysis of Electoral Advertising on Meta and Google during the EU 2024 Elections: Compliance, Transparency, and Targeting

Across 15 EU countries during the April-June 2024 EU Parliament election window, roughly 30,000 political ads were published and about €8.7 million spent on Meta and Google political advertising. Meta’s and Google’s own definitions of ‘political advertising’ were found not fully aligned with the EU’s Transparency and Targeting of Political Advertising Regulation (2024/900), with both platforms only partially meeting its transparency and targeting-disclosure requirements.

Interfering from Exile – Ilan Shor’s Demonstrations in Moldova

Exposed coordinated inauthentic behaviour promoting a Telegram bot tied to fugitive oligarch Ilan Shor ahead of Moldova’s 2025 election: a network of ~117 amplifying accounts republished his content to evade EU sanctions, while the bot offered $3,000/month to protest and harvested passport, geolocation and facial data from participants.

Polish Election Country Report 2025

A 28-organisation assessment of FIMI during Poland’s 2025 presidential elections, finding persistent foreign operations (Doppelganger, Operation Overload) and domestic amplification of anti-Ukrainian, anti-EU and anti-establishment narratives, whose impact was constrained by public resilience and active civil-society responses.

AI Generated Algorithmic Virality (AI Slop)

Across three European countries, 25% of TikTok's top search results contained synthetic AI imagery (80%+ from 'Agentic AI Accounts'); only about half of AI content was properly labeled despite DSA commitments.

App Stores & Meta's Porn: An Uncomfortable Affair

In Spring 2025 Meta approved 12,034+ pornographic ads, shown 12.8M times across all 27 EU countries, apparently violating store and Meta policies; the same images were auto-moderated when posted as normal Instagram posts.

Report on the 45th Canadian General Election

Monitoring of Canada's April 2025 federal election found election-themed scam ads widespread on Meta and X, CCP-aligned narratives circulating on Chinese-language platforms, and TikTok's feed favouring US content despite majority-Canadian creators; the report calls for stronger researcher data access.

From Sri Lanka to Canada: The Global Evolution of ‘Chameleon Ads’ on Meta’s Platforms

Writing independently, Sanjana Hattotuwa connected Social Media Lab’s Canadian ‘Chameleon ads’ investigation to earlier Sri Lankan waves of Facebook and Instagram crypto-scam advertising that impersonated journalists, ministers, the prime minister, the president and news outlets. He argues that the Canadian cases reflect tactics already developed in the Global South and highlights how disappearing Ad Library records obstruct comparative research.

The Hidden Game: How Scammers Use ‘Chameleon Ads’ to Bypass Meta’s Moderation

Social Media Lab identified and named ‘Chameleon ads’: Meta advertisers first obtain approval for harmless creative, then swap images, text or links for deepfake political endorsements and cryptocurrency scams, sometimes pausing and reverting the ad before review. The investigation also documented how non-political scam ads can disappear from Meta’s Ad Library without a durable public record.

Civil society files DSA complaint against Meta for toxic, profiling-fueled feeds

On 15 April 2025 EDRi, Bits of Freedom, GFF and Convocation Research + Design jointly filed a complaint with Ireland’s Coimisiún na Meán (Meta’s lead DSA regulator), alleging Meta’s deceptive interface design keeps Facebook and Instagram users on profiling-based feeds and silently resets any non-profiling choice, breaching DSA Articles 38, 27(3) and 25. On 5 May 2026, a year on, Coimisiún na Meán confirmed it had opened a formal investigation into Meta over these dark-pattern allegations.

Meta's Failing Ad Moderation: Health Scams Targeting EU Users

Identified 46,000+ ads promoting unapproved drugs and deceptive health claims, reaching EU users 292M+ times, violating Meta's own standards and potentially several DSA provisions.

The Dormant Danger: How Meta Ignores Large-Scale Inauthentic Behavior Networks of Malicious Advertisers

Mapped seven dormant networks totalling about 3.8 million Facebook pages reused for political propaganda and scams, and found Meta consistently removes individual ads while leaving the underlying inauthentic-behaviour networks intact, a systemic DSA compliance gap.

More Transparency and Less Spin: Analyzing Meta's policy changes

Meta's 2025 policy rollback could stop 97% of enforcement in key areas including hate speech.

Influence by Design: How Meta Accepted Russian Propaganda Payments Despite Sanctions

With Reset.tech and AI Forensics, revealed Meta ran 8,000+ pro-Russian ads tied to the sanctioned Social Design Agency, earning ~$338,000 between Aug 2023-Oct 2024 despite sanctions.

Statement on Meta's Announcement to Gut Moderation and Fact-checking

Argues Meta's end of third-party fact-checking and relocation of moderation teams will increase harm and disinformation; assesses the Commission's limited DSA options.

Pay-to-Play: Meta's Community (double) Standards on Pornographic Ads

3,000+ pornographic ads (including AI-generated media and celebrity deepfakes) were approved and generated 8M+ EU impressions, content that would be removed if posted by regular users.

The 2025 RDR Index: Big Tech Edition

Seventh edition (first full evaluation since 2022): no company scored over 50%; TikTok debuted strongly while X showed RDR's largest-ever single-company decline after cutting its human rights team and transparency reporting.

Toxic platforms, broken planet: Online abuse of environmental defenders

Of 204 land/environmental defenders surveyed, 92% had experienced online abuse or harassment tied to their work, with real-world safety and silencing effects.

Meta's Role in Romania's 2024 Presidential Election

With Reset Tech and EU DisinfoLab, examined Meta's advertising and its influence on public opinion during Romania's pivotal 2024 presidential vote.

Rainbow-burning: how social media companies increase risks for LGBTQ+ people in Africa

Documented 214 threatening anti-LGBTQ+ posts across the platforms in five African countries; only 51 were removed when reported, blaming under-resourced moderation for African languages and engagement-driven algorithms.

Doppelganger media impersonation: leveraging domain name dispute resolution

Explores using domain-name dispute resolution as a countermeasure against the Doppelganger media-impersonation operation.

So We Meet Again, Botiful: Latent Fake Ad Network Persists on Meta Two Years Since Doppelganger

Research note showing Meta continues to overlook large-scale networks of automated dormant fake pages reused for political and scam advertising, including Kremlin-affiliated Doppelganger campaigns targeting the EU and Moldova.

Independent Audit of Facebook's DSA Risk Assessment

Independent audit examining Facebook's systemic-risk assessment under the EU Digital Services Act (period 29 Aug 2023-30 June 2024).

Facebook's AI Failure

AI-generated images of fake Americans making political endorsements got millions of interactions before the US election.

TikTok and Facebook fail to detect harmful disinformation ads (US election)

Weeks before the US presidential election, TikTok approved 50% of disinformation ads and Facebook accepted one; YouTube approved 50% initially but blocked publication pending ID verification.

How Meta Brings in Millions Off Political Violence

A joint CalMatters/Markup investigation estimated Meta earns millions in ad revenue in the wake of violent news events, profiting from engagement surges around political violence.

Profiting from Propaganda: How Meta Benefits from Pro-Kremlin Election Meddling Ads in Moldova

Co-published with WatchDog.md: a pro-Kremlin campaign paid Meta about EUR 300,000 over 18 months, nearly a quarter of Moldova's political ad spend since August 2020, running 150+ inauthentic pages with disinformation and deepfakes ahead of Moldova's elections while Meta failed to dismantle the network.

US Election Platform Preparedness: Comparative Analysis

Assessed platforms across six areas ahead of the 2024 US election; the most alarming common finding was a pervasive lack of transparency about policies, safety efforts and resourcing.

Abusing Women in Politics: How Instagram is failing women

Instagram failed to act on 93% of reported abusive comments (including death/rape threats) targeting US women politicians.

Public Data Access Programs: A First Look

First systematic evaluation of 19 platforms' researcher data-access programs; access varies widely and data quality is poor.

Digital Apartheid in Gaza: Unjust Content Moderation (2-part series)

Documents how platforms, especially Meta (reportedly ~94% compliance with Israeli Cyber Unit takedown requests), disproportionately removed pro-Palestinian content while failing to curb anti-Palestinian hate.

Yet more evidence of Russia's boundless impunity to spread misinformation in the EU

Further documentation of the Doppelganger influence operation's continued reach across social platforms in the EU.

Meta's elections dashboard: A very disappointing sign

Technical critique of Meta's EU election-monitoring dashboard: keyword lists mis-copied across countries, Instagram lacking location filters, and Facebook filters cutting ~50% of relevant material, calling it a 'PR distraction.'

Unchecked Political Ads: Pro-Russian Propaganda on Meta Ahead of EU Elections

Between May 1-27 2024, Meta approved at least 275 pro-Russian propaganda ads (lacking required disclaimers) targeting Italy, Germany, France and Poland, reaching over 3 million accounts.

Antisemitic Admins: Meta's failure to moderate large Facebook Groups

Meta failed to act on admins spreading anti-Jewish hate to 300K+ Facebook Group members.

Confusing Google predictions and unbalanced politics on TikTok and Instagram

CrossOver Finland analysis found politically unbalanced content on TikTok and Instagram alongside confusing Google search predictions ahead of the 2024 elections.

Assessing cost-effectiveness: responses to the Doppelganger operation

Assesses responses to the Russian Doppelganger operation that clones media outlets and spreads links/ads across X, Facebook and Telegram.

Impact of regulation on children’s digital lives

Meta, Google, TikTok and Snap made 128 documented child-safety and privacy changes between 2017 and 2024, peaking at 42 changes in 2021 when the UK Age Appropriate Design Code took effect; Meta alone accounted for 61 of the changes. Only 8 of 50 companies contacted with information requests responded.

Shrinking democracy, growing violence: Internet shutdowns in 2023

Annual #KeepItOn documentation of shutdowns and platform blocking worldwide in 2023, tying disruptions to elections, protests and conflict.

No Embargo In Sight: Meta Lets Pro-Russia Propaganda Ads Flood the EU

A moderation loophole let pro-Russian propaganda run on Meta; under 5% of undeclared political ads were caught, and one campaign reached 38M+ users in France and Germany. The European Commission opened proceedings citing this report.

Full Disclosure: Stress Testing Tech Platforms' Ad Repositories

With CheckFirst: tested ad-transparency tools of 11 companies; none rated 'ready,' with major data gaps and inaccuracies.

Mozilla commissions CheckFirst to conduct a stress test of Ad Repositories

First comprehensive report on DSA-strengthened platform ad libraries, finding the ad transparency tools failing to meet their intended purpose.

Digital Disparities: the global battle for reproductive rights on social media

Meta restricts reproductive-care information in Africa/Asia/Latin America while allowing anti-choice misinformation ads.

Instagram Creators: Check If Your Posts Are 'Political'

After Meta stopped 'proactively recommending' political content, The Markup built a tool letting creators test whether their posts would be classified as political and thus algorithmically limited.

How Automated Content Moderation Works (Even When It Doesn't)

Explainer documenting the automated techniques (hash-matching, classifiers) platforms use to moderate billions of posts, and where these systems routinely fail.

Safe by Default (People vs Big Tech briefing)

Argues engagement-based recommender systems amplifying borderline content is a feature not a bug; calls on regulators to require non-profiling feeds as the default.

Platforms, Promises and Politics: a reality check on election pledges

Finds platforms' election interventions in the Global Majority are template-copied and far weaker than those offered in the US/EU.

Demoted, Deleted, and Denied: More Than Just Shadowbanning on Instagram

An investigation found Instagram heavily demoted nongraphic Israel-Hamas war images, deleted captions, hid comments without notification, and suppressed hashtags, well beyond classic 'shadowbanning.'

It's not a glitch: how Meta systematically censors Palestinian voices

Documents systematic censorship of Palestinian and pro-Palestinian content since October 7, 2023, driven by Meta's vague 'Dangerous Organizations' policy and discriminatory enforcement (e.g., lowering the hide-threshold to 25%).

Each Facebook User is Monitored by Thousands of Companies

A study of 267,000 users found the average Facebook user had activity sent to Facebook by roughly 2,230 different companies, exposing the scale of third-party tracking.

Disinformation on Facebook: research and content moderation policies

Factsheet documenting how disinformation on Facebook is researched and how the platform's content-moderation policies operate.

Algorithms of Trauma 2: 'Anxious about your health? Facebook won't let you forget'

Case study showing Facebook's ad-delivery algorithms push distressing health/illness ads exploiting users' mental vulnerabilities; disabling 'sensitive interests' limits advertiser targeting but not Facebook's own profiling.

Violent and sexualised hate speech targeting women journalists (South Africa)

Nearly all 40 misogynistic test ads (in 4 languages) were approved; Meta and TikTok approved all 40 within 24 hours. Concludes automated moderation is 'not fit for purpose.'

Facebook Watches Teens Online As They Prep for College

A Pixel Hunt investigation found Meta's tracking pixel embedded on school, test-prep and college-application websites, letting Meta monitor students as they browse education sites.

Not a solution: Meta's new AI system to contain discriminatory ads

Meta's 'Variance Reduction System' for biased US housing-ad delivery is 'more band-aid than AI fairness solution'; reporting gaps make it impossible to verify.

Under the Radar: Vast Networks of Fake Accounts Raise Questions About Meta's Compliance with the EU's New Digital Rulebook

Uncovered an ad network of at least 242,000 fake Facebook pages, alongside a second ecosystem of more than 340,000 pages, pushing Russian propaganda and scams across 32+ EU countries; despite prior warnings from September 2022, Meta had failed to act, making this one of the earliest and most-cited DSA compliance challenges to Meta.

Platforms' policies on climate change misinformation (V1)

Maps how the five platforms address climate misinformation, revealing uneven and weak climate-specific rules.

A world of online hate and lies: Mapping our investigations

A synthesis of 10+ country investigations concluding platforms systematically fail to enforce their own hate-speech/disinformation ad policies, with weaker moderation outside the US.

Social media algorithms are harmless, or are they?

Critiques the Meta-partnered 2020-election studies, arguing behavioral data showed the algorithms increased uncivil-content exposure, removed cross-cutting content, and boosted time-on-platform.

Facebook Hustles: The Belgian spin-off

Deep-dive into a scam network of 1,500+ Facebook ads luring users to fake media sites impersonating Belgian outlets and politicians.

Platforms' policies on elections misinformation

Maps and compares how five major platforms address election misinformation in their policies.

Platforms' policies on health misinformation

Maps and compares how five major platforms address health misinformation in their policies.

Facebook Hustles: The Hidden Mechanics of a Scam Machinery

A month-long investigation uncovered a large-scale scam operation recruiting victims via 1,500+ Facebook ads and fake media sites impersonating news outlets/creators.

'We need to kill them': Xenophobic hate speech approved by Facebook, TikTok and YouTube (South Africa)

Violent xenophobic test ads were approved by all three platforms; nearly every ad passed, with only one Facebook rejection.

Reels of Fortune: Instagram-shaped memories for a bigger reach

Qualitative story on how Instagram's pivot to Reels/video pushes users to curate memories 'for the algorithm,' generating training data that teaches the platform to rank memories automatically.

Horizon Worlds Exposed

Minors were routinely harassed and exposed to adult content in Meta's flagship VR social network.

Extreme and violent anti-LGBTQ+ hate approved for publication by leading social media platforms

Test ads with extreme anti-LGBTQ+ hate were almost all approved: YouTube and TikTok approved all 10 ads, Facebook rejected only two, and all three approved the 'burn all gays' ad.

Pathways: How digital design puts children at risk

Using avatar (test) accounts posing as 13-year-olds, researchers found children were directly targeted with harmful content — including material promoting eating disorders, extreme diets, self-harm and suicide — within 24 hours of creating an account on Facebook, Instagram or TikTok. Child accounts also received contact from unknown adults directing them toward further harmful content.