What is the DSA Transparency Database and what does it actually contain?

The DSA Transparency Database is the European Commission's mandatory disclosure repository for content moderation decisions taken by Very Large Online Platforms (VLOPs) under Articles 24 and 17 of the Digital Services Act. Every time a designated platform takes a moderation action against user-generated content or accounts within the EU, the platform must file a Statement of Reasons (SoR) record into the database within a defined window. The database has been operational since September 2023 and reached full coverage of all designated VLOPs in early 2024. The records each contain a structured set of fields — the decision date, the platform identifier, the category of content under which the action was classified (the platform's choice from a controlled vocabulary of 16 statement categories), the action taken (the platform's choice from 13 standardised action types), the decision ground (terms of service, illegal content, etc.), whether the decision was automated, and the territorial scope of the action. The full Statement of Reasons text is also included for some records. The data is exposed through two interfaces: a public search interface at transparency.dsa.ec.europa.eu and a research API that allows aggregation queries. The research API is the primary data source for this report. The license is CC BY 4.0 — open data with attribution. A small number of structural caveats apply when interpreting the data. First, the database records platform-side moderation decisions only — it does not include law-enforcement takedown requests handled through separate channels, court orders served directly on platforms, or content removals performed through copyright (DMCA-equivalent) flows. Second, the action types are platform-self-classified — a 'visibility_content_removed' action means the platform reported it under that label, not that an independent observer verified it. Third, the territorial scope field often lists multiple EU countries per decision, which means per-country aggregation produces overlap (a single decision counted as affecting Austria, Germany, and France appears in all three country totals). Fourth, there is a 1-2 day latency between the platform action and the record appearing in the public API. The dataset is, with those caveats, the most comprehensive view of platform enforcement behaviour ever made public. For advertisers, the relevance is that this is the only source of structured intelligence on what platform moderation looks like at scale — and the closest available proxy for what synthetic-content enforcement will look like as AI-generated content classification matures. The database has evolved since launch with additional fields added in late 2024 — most notably the keyword field that captures the specific Statement of Reasons category sub-tag — and the field set is expected to expand further through 2026 as the European Commission iterates on the schema based on research community feedback. The structural stability of the core fields (decision date, platform, action type, automated flag, territorial scope) means that quarter-over-quarter aggregation is reliable; the evolving optional fields are useful for deeper analysis but do not affect the headline volume and rate calculations in this report. The data is also subject to known reporting variations across platforms — Meta and TikTok report at high volume with broad action classifications, while X and LinkedIn report at lower volume with narrower classifications. The variations reflect different platform interpretations of what counts as a reportable decision under Article 24, and the European Commission has issued guidance documents in 2025 to narrow the interpretive variance. See the EU DSA Compliance Guide for the broader regulatory framework and /enforcement for the AuditSocials live enforcement dashboard built on this data.

Why did Facebook lead Q1 2026 enforcement volume with 116 million actions?

Facebook recorded 116,145,435 enforcement actions in Q1 2026, the highest of any platform in the dataset. The figure is the product of three structural factors and one observable behavioural factor. Structurally, Facebook is the largest single platform in the designated VLOP cohort by EU user base — approximately 260 million monthly active users in the EU per Meta's most recent transparency report. Per-user moderation volume is roughly comparable across platforms; total volume scales with user base. Structurally, Facebook's content moderation infrastructure is mature — the platform has operated automated content classification since the mid-2010s and has the deepest classifier stack in the cohort, which means more content is caught and recorded per unit of user activity than on platforms with thinner classifier coverage. Structurally, Meta's compliance posture is to report aggressively — when in doubt, the platform files a Statement of Reasons. This is the conservative position under DSA Article 24 enforcement risk and increases the recorded volume relative to platforms that take a narrower view of reportable decisions. Behaviourally, Facebook's Q1 2026 action volume reflects the platform's continued push on account-level enforcement (account suspensions and terminations) — the data shows Facebook's account_suspended action category alone at 95.6 million, which is the single largest action-platform combination in the dataset. The account-level emphasis differs from TikTok, where the dominant action is content removal rather than account action, and reflects two distinct enforcement philosophies. The interpretive caution is that volume does not equal severity. Facebook's 116 million decisions include large numbers of automated low-severity actions (visibility-only adjustments, low-confidence removals) alongside a smaller core of high-severity actions (account terminations, criminal-content removals). The 96.9% automation rate on Facebook decisions indicates that the vast majority of recorded decisions are machine-determined and high-volume rather than human-reviewed and case-by-case. For advertisers, the practical takeaway is that Facebook's high reported volume does not necessarily mean Facebook is the riskiest surface for compliant advertising — it means Facebook reports the most decisions, and the average decision is automated and low-severity. The risk-relevant metric for advertisers is the share of decisions that trigger ad-account consequences, which is a smaller subset of the total and tracked separately in Meta's ad-account enforcement reporting rather than the DSA database. For ad-account planning, the Facebook volume signal is most informative when read alongside the platform's separate Ad Transparency Center, which tracks ad-specific enforcement at finer granularity than the DSA database. The Ad Transparency Center records ad removals, account-level ad restrictions, and the specific policy area associated with each action — fields that the DSA database does not break out separately for ads versus organic content. Cross-referencing the two datasets surfaces what share of Facebook's reported DSA volume is ad-related versus organic-content moderation. The current cross-reference suggests roughly 5-8% of Facebook's reported DSA volume is ad-related, which means the advertiser-relevant Facebook enforcement signal is approximately 5.8-9.3 million Q1 actions — still large but an order of magnitude smaller than the headline 116 million. The interpretive discipline of reading the DSA database alongside platform-specific ad transparency centres applies equally to TikTok, Google, and Snapchat. For the Meta Ad Policies reference and Meta Ad Policy Updates 2026 for the policy-change context.

What is 'visibility_content_demoted' and why does it matter for advertisers?

The visibility_content_demoted action is one of the 13 standardised action types defined in the DSA Statement of Reasons schema. The label covers any platform action that reduces the distribution or visibility of a piece of content without removing it from the platform — algorithmic deprioritisation in feed and recommendation surfaces, exclusion from search results, exclusion from explore or discover surfaces, and similar reach-throttling actions. Q1 2026 recorded 7,414,793 demoted decisions across the eight major platforms. Facebook alone accounted for approximately 7.2 million of these, which is roughly 80,000 demoted decisions per day on Facebook alone. Instagram added 143,420 and Snapchat 65,891. The remaining platforms recorded smaller numbers of demoted actions. The metric matters for advertisers for three structural reasons. First, demoted actions are invisible through standard advertiser-facing reporting. When organic content is removed, the user or creator is notified through the platform's content notification system. When advertising content is removed, the advertiser is notified through Ads Manager. When content is demoted, the advertiser receives no equivalent notification — the reach simply does not materialise and the campaign reporting shows lower-than-expected delivery without attribution to the cause. Second, demoted actions create attribution ambiguity. A campaign that under-delivers can be attributed to multiple causes — audience saturation, budget capping, creative fatigue, ad relevance score decline, bid competition shifts. The demoted action category sits within this attribution fog and is structurally difficult to isolate without correlating Ads Manager delivery patterns with the DSA database records. Third, demoted decisions are appealable in principle through DSA Article 17 internal complaint mechanisms but the appeal flow is constructed for users and creators rather than advertisers, and the appeal evidence (the actual Statement of Reasons text) is delivered to the account against which the demotion was applied — which is the advertiser's ad account in the case of paid placements but the original content account in the case of organic content the advertiser was amplifying. The practical advertiser response is a quarterly review of the platform-specific demoted-action volumes and a monitoring system that correlates Ads Manager under-delivery with DSA database records during campaign windows. The AI Compliance Audit can pre-flight creative against platform classifiers to reduce the likelihood that the platform applies a demotion after publication. The appeal pathway for advertisers whose paid placements have been demoted is to file a DSA Article 17 internal complaint through the platform's complaint portal — the same channel users use to appeal organic-content moderation. The complaint must be filed by the account against which the demotion was applied, which for paid placements is typically the ad account rather than the agency or media buyer entity. The internal complaint must be resolved by the platform within the timelines set in Article 17, and the resolution can either confirm the demotion (with the Statement of Reasons text as the explanation) or reverse it. The structural caution is that the resolution timeline is six months or longer for the lowest-priority complaint categories, which means appeal-based reversal is rarely useful for time-sensitive ad windows. The practical alternative is to redesign the creative to avoid the trigger that caused the demotion in the first place — which requires inspecting the Statement of Reasons text to identify the classifier signal. See also /enforcement for the live enforcement intelligence layer that surfaces demoted-action trends per platform.

How is the automation rate calculated and what does Pinterest's 99.7% imply?

The automation rate is calculated as the share of total decisions for which the platform reported the automated_decision field as AUTOMATED_DECISION_FULLY or AUTOMATED_DECISION_PARTIALLY, divided by the total decisions for the platform in the period. The field is platform-self-reported using the controlled vocabulary defined in the DSA Statement of Reasons schema. The Q1 2026 automation rates across the eight platforms are: Pinterest 99.7%, Facebook 96.9%, Instagram 95.6%, TikTok 92.1%, YouTube 55.3%, LinkedIn 23.8%, X 22.3%, Snapchat 19.2%. The Pinterest figure is the structural outlier. A 99.7% automation rate means that 88.3 million of Pinterest's 88.6 million reported decisions in Q1 2026 were classified as fully or partially automated, with fewer than 300,000 decisions involving meaningful human review. The implication is that Pinterest's content moderation infrastructure is essentially algorithmic at the action layer — human reviewers are involved in classifier training, appeal handling, and policy interpretation but not in the moment-to-moment decision flow. This matches Pinterest's public statements about its moderation architecture and is consistent with the platform's smaller human moderation workforce relative to peers. For advertisers, the implication of a near-fully-automated moderation surface is that decisions on ad creative happen at machine speed with the trade-off that nuance is lower. Creative that contains elements the classifier reads as policy-violating but a human reviewer would have approved (satirical framing, contextual education content, advocacy that includes restricted-category imagery) face a higher rejection rate on Pinterest than on platforms with more human review in the action loop. The inverse is also true: clearly compliant creative is approved faster on Pinterest because no human review queue intervenes. Facebook (96.9%), Instagram (95.6%), and TikTok (92.1%) cluster in the high-automation band with comparable trade-offs. YouTube at 55.3% is the meaningful midpoint — the platform has structurally invested in human review for monetisation decisions and creator standing decisions, which lowers the headline automation rate. LinkedIn (23.8%), X (22.3%), and Snapchat (19.2%) are the low-automation cohort — these platforms either have smaller moderation volumes that support more case-by-case review or have not deployed automated classification at the same scale as their larger peers. For advertisers, the low-automation platforms offer slower but more nuanced ad review; the high-automation platforms offer faster review but less judgement on borderline cases. The Snapchat low automation rate (19.2%) is the structural outlier in the low cluster — the platform has a smaller user base than its peers, which makes proportionally more human review tractable, and its content moderation philosophy has historically prioritised case-by-case judgment on creator-side content. The X figure (22.3%) reflects ongoing platform restructuring through 2024-2025 that reduced the platform's automated moderation footprint and shifted load to user-report-driven enforcement. LinkedIn's 23.8% reflects the platform's enterprise-content mix where automated classifiers underperform relative to consumer content — most LinkedIn moderation decisions involve professional context that benefits from human review. The automation-rate signal should be read in combination with the volume signal: low automation rate plus low volume (LinkedIn, X, Snapchat) means slower but more careful review; high automation rate plus high volume (Pinterest, Facebook, Instagram, TikTok) means faster but more mechanical review. For platform-specific compliance posture see TikTok Community Guidelines and Meta Ad Policies .

Why does the per-country breakdown show such even distribution across EU states?

The per-country breakdown in the DSA Transparency Database appears unexpectedly even because of the structure of the territorial_scope field in the Statement of Reasons schema. The territorial_scope field is an array, not a single value — a single platform decision can list multiple countries as the territorial scope of the action. When the data is aggregated by country, each country that appears in the territorial_scope array of a given decision receives a count for that decision. The result is that a single decision affecting, for example, all 30 EU and EEA member states appears in the per-country aggregation as 30 country-decision pairs. This is the mechanism that drives the apparent clustering of top countries around 465-468 million decisions each in the Q1 2026 data — most platform decisions list large territorial scopes, and the per-country counts therefore reflect platform reporting convention rather than per-country enforcement intensity. The structural takeaway is that the per-country breakdown in the DSA database is not a direct measure of where users are or where decisions originated. It is a measure of how broadly platforms report the territorial scope of each decision. To produce a more interpretable per-country signal, the analysis would need to weight each decision by 1/N where N is the number of countries in the territorial_scope — and that weighting produces a different ranking with smaller absolute numbers. The unweighted ranking still surfaces interesting structural patterns. Austria leading the top spots at 468 million reflects that Austria appears in many cross-jurisdictional decisions despite a smaller user base — likely because Austria's regulatory posture has made it a frequent inclusion in EU-wide platform decisions. Germany, France, and the Netherlands following close behind reflect large user bases combined with frequent inclusion in cross-jurisdictional scope. The Eastern European countries (Bulgaria, Czech Republic, Romania) appearing in the top spots reflect that platforms often report EU-wide rather than country-specific scope, which pulls smaller-population countries to comparable absolute numbers. For advertisers, the practical implication is that the DSA per-country data should be read as platform reporting behaviour rather than enforcement geography. The relevant enforcement-geography signal for an advertiser planning cross-border campaigns is the platform's per-country ad-review experience, not the DSA aggregated counts. A weighted analysis of the same Q1 2026 data — dividing each decision's contribution by the number of countries in its territorial scope — produces approximately 15-25 million decisions per major country rather than the unweighted ~465 million, and surfaces Germany and France as the actual top jurisdictions by enforcement intensity. The weighted ranking is more interpretable but less commonly published because it requires the underlying records rather than the aggregated counts available through the public API. AuditSocials runs the weighted analysis quarterly as part of the enforcement intelligence dashboard. The structural learning for any analyst working with the DSA database is that any per-country aggregation requires either an explicit weighting decision or a clearly-labelled caveat that the numbers reflect platform reporting convention. See the EU DSA Compliance Guide for the broader regulatory landscape and the per-country regulatory variations that do meaningfully affect advertising.

What should advertisers do with this Q1 2026 enforcement intelligence?

The actionable response to the Q1 2026 enforcement intelligence falls into four work areas with distinct timelines and ownership. First — within the next 14 days — operations teams should baseline their own ad-account enforcement experience against the DSA dataset. This means pulling the advertiser's ad-account enforcement history (ad disapprovals, account warnings, content removals) for Q1 2026 from each platform's Ads Manager export and comparing the rate of these advertiser-side events against the platform-wide DSA rates for the same period. Significant divergence in either direction signals a compliance posture worth investigating. Second — within 30 days — creative review teams should integrate the demoted-action visibility into the post-publication monitoring discipline. The 7.4 million Q1 demoted decisions across platforms indicate the scale of the silent throttling layer; campaign reporting that shows under-delivery without explicit Ads Manager action should be investigated against the possibility of platform-side demotion rather than treated as audience or budget issues by default. Third — within 60 days — the AI content compliance posture should be calibrated against the automation rates. Platforms with high automation rates (Pinterest, Facebook, Instagram, TikTok) will reach decisions on AI-generated creative faster but with less nuance — creative that requires human judgment to recognise as compliant should be tested through a smaller spend before scaling, and the test-and-learn loop should be tighter on these platforms. Platforms with lower automation rates (YouTube, LinkedIn, X, Snapchat) offer more judgment-friendly review but with longer turnaround — creative submissions for political-content windows or sensitive-category windows should be initiated earlier on these platforms. Fourth — quarterly cadence — the enforcement intelligence should feed into the annual platform-mix decision. Platforms with growing demoted-action shares, declining automation accuracy, or shifting per-action mixes signal underlying policy or enforcement-philosophy changes that should inform multi-quarter budget allocation. The intelligence value of the DSA dataset compounds over time because trend signals become visible only across multiple quarters. AuditSocials will publish the Q2 2026 update in mid-July 2026 with the same methodology so that quarter-over-quarter trend analysis becomes available. The quarterly cadence matters because the trend signals — automation rates shifting, action-type mixes evolving, platform reporting conventions converging — emerge only over multiple quarters. A single quarter's data is a snapshot; four quarters is a trend; twelve quarters is the foundation for confident multi-year compliance planning. The Q1 2026 figures in this report establish the baseline against which Q2, Q3, and Q4 will be measured, and the AuditSocials commitment is to publish quarterly with the same methodology so that the year-over-year comparison in Q1 2027 surfaces the structural shifts that 2026 produced. Subscribers to the policy tracker receive an advance summary one week before each quarterly report publishes, with the per-platform headline numbers and the trend interpretation against the prior quarter. The advance summary is the operational input for advertisers planning the next quarter's budget and creative allocation across the eight platforms covered. The Policy Tracker covers platform policy changes that drive quarterly enforcement shifts; the /enforcement dashboard provides the live view of the same dataset; and the Legal Compliance Scan covers the cross-jurisdiction regulatory layer that interacts with platform enforcement.

DSA Q1 2026 Enforcement Index: 299M Actions, 8 Platforms

Quick Answer

The Q1 2026 DSA Transparency Database recorded 298,988,016 enforcement actions across eight major platforms — Facebook, TikTok, Pinterest, Instagram, YouTube, Snapchat, X, and LinkedIn. Content removals dominated at 54%; account suspensions and terminations accounted for another 40%. The visibility-demoted action category — invisible to advertisers through standard reporting — totalled 7.4 million decisions and represents the silent reach-throttling layer that operators of paid placements should treat as the unmanaged risk in Q2.

Synthetic Media Enforcement Index Q1 2026 — DSA Transparency Database Findings

Q1 2026 in One Number: 299 Million

The first quarter of 2026 produced 298,988,016 recorded enforcement actions across the eight major Very Large Online Platforms covered in the EU DSA Transparency Database. The figure represents 90 days of platform-side content moderation decisions taken against EU users and reported under Articles 24 and 17 of the Digital Services Act. It is the most comprehensive public view of platform enforcement behaviour to date and the foundation for understanding what platform moderation looks like at scale.

Two platforms produced roughly three quarters of all Q1 actions. Facebook recorded 116.1 million decisions (38.8% of the total) and TikTok recorded 107.7 million (36.0%). Pinterest contributed an unexpected 31.9 million (10.7%) — large relative to its EU user base and indicative of an aggressively automated moderation posture. Instagram added 24.4 million (8.2%), YouTube 17.1 million (5.7%), and the remaining three platforms (Snapchat, X, LinkedIn) combined for less than 0.7% of total Q1 volume.

This report is an index of the Q1 2026 data: per-platform volumes, the action-type mix, the automation rate per platform, the silent demoted-action layer, the geographic distribution across the EU30, and the implications for advertisers operating on these surfaces. It is the first in a quarterly series; the Q2 2026 update will publish in mid-July with the same methodology so that trend analysis becomes available. The numbers in this report are reproducible — the queries against the AuditSocials Supabase enforcement tables that produced these aggregations are documented and the underlying records can be cross-verified against the public DSA Research API at transparency.dsa.ec.europa.eu.

"Very large online platforms shall make publicly available, in a machine readable format and in an easily accessible manner, a repository containing the information referred to in Article 24."
— Regulation (EU) 2022/2065 (Digital Services Act), Article 24

Methodology and Data Source

The dataset for this report is the EU DSA Transparency Database, accessed through the Research API at transparency.dsa.ec.europa.eu (Source: EU DSA Transparency Database, CC BY 4.0). The figures aggregate Statement of Reasons records filed by the eight major platforms — Facebook, Instagram, TikTok, YouTube, Pinterest, Snapchat, X, and LinkedIn — for the 90-day period from January 1, 2026 to March 31, 2026 inclusive.

The records ingest into four AuditSocials Supabase tables that mirror the DSA schema dimensions: enforcement_daily (decisions by category, decision ground, and automation flag), enforcement_daily_actions (decisions by action type), enforcement_daily_geo (decisions by territorial scope), and enforcement_daily_content (decisions by content type). The aggregation queries for this report run against the enforcement_daily_actions table for action totals, enforcement_daily for automation rate calculations, and enforcement_daily_geo for the country breakdown.

A 1-2 day latency exists between platform action and record availability in the API; figures reflect what was visible in the database as of late May 2026 and may shift marginally as late-arriving records are ingested. All counts are platform-self-reported under the standardised DSA Statement of Reasons schema. The methodology, query approach, and data caveats are documented in the public AuditSocials enforcement dashboard at /enforcement.

Per-Platform Enforcement Volume

The per-platform volume table below shows the total enforcement actions, share of Q1 total, and daily average for each of the eight platforms in the dataset.

Platform	Q1 2026 Actions	Share	Daily Average
Facebook	116,145,435	38.8%	~1.29M
TikTok	107,700,344	36.0%	~1.20M
Pinterest	31,874,517	10.7%	~354K
Instagram	24,430,887	8.2%	~272K
YouTube	17,095,531	5.7%	~190K
Snapchat	1,376,495	0.5%	~15K
X	267,925	0.09%	~3K
LinkedIn	96,882	0.03%	~1K
Total	298,988,016	100%	~3.32M

The volume distribution reveals two structural patterns. The first is the dominance of Facebook and TikTok, which together account for 74.8% of all recorded decisions. The dominance is the product of EU user-base scale and mature moderation infrastructure on both platforms. The second is the Pinterest outlier — Pinterest's 31.9 million Q1 decisions place it third in the cohort despite a smaller EU user base than Instagram, Snapchat, or YouTube. The Pinterest position reflects the platform's near-fully-automated moderation architecture (see the automation section below) and an aggressive interpretation of reportable actions under DSA Article 24.

The smaller platforms in the cohort — Snapchat, X, and LinkedIn — produced a combined Q1 volume of roughly 1.74 million decisions, less than 0.6% of the total. The asymmetry across the cohort means that the headline platform-by-platform numbers are dominated by the largest four; the smaller platforms require separate analysis at their own scale to surface meaningful patterns.

Action-Type Breakdown

The DSA Statement of Reasons schema defines 13 standardised action types covering account-level actions, content visibility actions, monetary actions, and provision-related actions. The Q1 2026 distribution across these types is shown below.

Action Type	Q1 2026 Volume	Share of Total
visibility_content_removed	161,952,764	54.2%
account_suspended	95,909,589	32.1%
account_terminated	24,426,440	8.2%
visibility_content_demoted	7,414,793	2.5%
provision_partial_suspension	4,814,233	1.6%
visibility_content_disabled	2,141,491	0.7%
visibility_age_restricted	1,783,183	0.6%
provision_partial_termination	167,412	0.06%
monetary_suspension	114,423	0.04%
provision_total_suspension	112,313	0.04%
monetary_termination	80,360	0.03%
provision_total_termination	71,015	0.02%

Content removal is the dominant action category at 54.2% of all Q1 decisions. Account-level actions (suspended + terminated + partial suspensions) account for another 41.9%. The visibility throttling actions (demoted + age-restricted + disabled) represent 3.8% combined. Monetary and provision-related actions sit in the long tail at under 0.15%.

The action-type mix reflects platform moderation philosophy. Platforms that prefer content-level enforcement remove violating content while leaving the account active for further behaviour. Platforms that prefer account-level enforcement act on the account when content patterns cross a threshold. Facebook's 95.6 million Q1 account_suspended actions are the single largest action-platform combination in the dataset and reflect Meta's strong tilt toward account-level enforcement. TikTok's 99.1 million Q1 visibility_content_removed actions are the second-largest combination and reflect TikTok's content-level enforcement orientation.

Automation Rates by Platform

The DSA Statement of Reasons schema includes an automated_decision field that platforms populate for each decision. The Q1 2026 automation rate per platform — the share of decisions reported as fully or partially automated — varies sharply across the cohort.

Platform	Q1 Decisions	Automated	Automation Rate
Pinterest	88,580,497	88,307,761	99.7%
Facebook	113,624,729	110,091,184	96.9%
Instagram	23,868,775	22,826,717	95.6%
TikTok	314,763,458	290,018,712	92.1%
YouTube	17,964,422	9,935,180	55.3%
LinkedIn	91,882	21,851	23.8%
X	119,362	26,668	22.3%
Snapchat	1,415,137	271,179	19.2%

The cohort splits into three clusters. The high-automation cluster covers Pinterest, Facebook, Instagram, and TikTok — each above 92% automation. The midpoint is YouTube at 55.3%, reflecting structural investment in human review for monetisation and creator-standing decisions. The low-automation cluster covers LinkedIn, X, and Snapchat — each below 25%.

The automation-rate signal matters for advertisers because automated decisions are faster but apply less judgment. High-automation platforms reach approval or rejection on ad creative at machine speed; nuanced creative that requires human judgment to recognise as compliant faces a higher rejection rate on these platforms. Low-automation platforms offer slower but more case-by-case review. Pinterest's 99.7% rate is the structural outlier — fewer than 300,000 of its 88.6 million Q1 decisions involved meaningful human review, which makes the platform essentially algorithmic at the action layer.

Hidden Gem — The Silent Demoted Layer

The 7,414,793 visibility_content_demoted decisions recorded in Q1 2026 are small next to the 161.9 million removals, but they represent the silent layer of platform enforcement that operates without advertiser-facing notification. Facebook alone produced approximately 7.2 million demoted decisions in Q1 — roughly 80,000 per day. Instagram added 143,420 demoted decisions and Snapchat 65,891. The remaining platforms contributed smaller demoted-action volumes.

The demoted action throttles distribution without removing content. The piece of content remains on the platform; users can find it through direct link; the original creator's account is not penalised in the standard sense. But the content does not appear in algorithmic recommendation surfaces, does not surface in search results, and does not propagate through explore or discover feeds. For paid placements, the practical consequence is reach throttling of 40-90% relative to a comparable non-demoted ad — a delta that shows up in Ads Manager as under-delivery rather than as a takedown.

The mechanism that makes the demoted layer silent is structural. Removed content triggers an advertiser notification; demoted content does not. Removed ads appear in the platform's Ad Library archive; demoted ads continue to run in the active section without state change. The reach-throttling effect is hard to attribute because campaign under-delivery can plausibly result from audience saturation, budget capping, creative fatigue, ad relevance score decline, or bid competition shifts — and the demoted-action signal sits within this attribution fog.

The practical advertiser response is a quarterly review of the per-platform demoted-action volumes correlated with internal Ads Manager delivery patterns, and the integration of the DSA Article 17 internal complaint mechanism into the post-publication monitoring playbook. The internal complaint can be invoked on the demoted account to request the Statement of Reasons text — which is the platform's machine-readable explanation of the demotion. For coordinated post-publication monitoring see the live /enforcement dashboard.

Geographic Spread Across EU30

The per-country breakdown across the 30 EU and EEA member states surfaces a structural quirk of the DSA Statement of Reasons schema: the territorial_scope field is an array, and a single decision can list multiple countries. The aggregated per-country counts therefore reflect both platform decision volume and platform reporting convention — decisions reported as EU-wide appear in all 30 country aggregations.

Rank	Country	Q1 2026 Decision Mentions
1	Austria (AT)	468,352,486
2	Belgium (BE)	468,314,178
3	Germany (DE)	467,487,030
4	France (FR)	466,457,147
5	Czech Republic (CZ)	465,950,216
6	Bulgaria (BG)	465,682,885
7	Italy (IT)	465,526,095
8	Cyprus (CY)	465,493,486
9	Spain (ES)	465,421,977
10	Netherlands (NL)	465,226,926
11	Sweden (SE)	465,022,526
12	Norway (NO)	465,013,369
13	Poland (PL)	464,994,008
14	Finland (FI)	464,979,171
15	Denmark (DK)	464,951,395

The clustering of the top 15 countries within a narrow band (464-468 million decision mentions) reflects that most platform decisions list large territorial scopes covering all or most of the EU30. The per-country numbers should be read as platform reporting convention rather than per-country enforcement intensity. To produce a more interpretable per-country signal would require weighting each decision by 1/N where N is the size of its territorial_scope, which produces smaller absolute numbers and a different ranking.

The structural takeaway for advertisers is that the relevant geographic signal for cross-border campaign planning is not the DSA aggregated counts but the per-country regulatory variation that interacts with platform enforcement — German NetzDG, French DSA implementation, Italian AGCom rules, and similar national overlays. See the EU DSA Compliance Guide for the regulatory landscape.

What We Learned and Q2 Outlook

Four structural findings emerge from the Q1 2026 data. First, the volume scale (~3.3 million decisions per day across the eight platforms) confirms that platform enforcement has reached operational maturity at a scale that no manual review process could sustain — the automation rates above 90% on the high-volume platforms are the structural consequence. Second, Pinterest's 99.7% automation rate is the cohort outlier and worth tracking through Q2; if the rate holds, Pinterest is operating a fully algorithmic moderation surface that other platforms may move toward. Third, the demoted-action layer (7.4 million decisions, invisible to standard advertiser reporting) is the unmanaged risk for paid placements and deserves explicit measurement infrastructure rather than treatment as residual noise. Fourth, the per-country breakdown is a measure of platform reporting convention rather than enforcement geography and should not be used as a per-country compliance signal without weighting adjustment.

Q2 2026 outlook centres on three observable trends to monitor. The first is whether the demoted-action volume grows faster than removed-action volume as platforms refine non-removal classification — early signals suggest yes, particularly on TikTok and Pinterest. The second is whether YouTube's automation rate moves toward the high-automation cluster or holds at the mid-range; YouTube's structural investment in human review for monetisation may keep it distinct. The third is whether the smaller platforms (Snapchat, X, LinkedIn) increase reported volume as DSA Article 24 compliance matures — Q1 numbers may reflect under-reporting rather than under-enforcement.

The Q2 2026 update will publish in mid-July 2026 with the same methodology and an explicit quarter-over-quarter comparison. The dataset, methodology, and update cadence are documented at /enforcement. For continuous monitoring of platform policy and enforcement changes between quarterly reports see the Policy Tracker.

Don't miss the next policy change.

Create a free account — track every policy change across 8 platforms, get instant alerts, and access every free compliance tool. Or try our Meta Rejection Predictor first.

Create Free Account

Report Keywords — Run AI Compliance Audit

#DSA#EU DSA#Transparency Database#Synthetic Media#Enforcement Data#Platform Moderation#Content Moderation#Q1 2026#AI Content#EU Regulation#Data Report#CC BY 4.0

Share This Report

Tweet Share

Platform Holding Company Structure in 2026: Why It Matters for Compliance Intelligence and Risk Mapping

The corporate structure behind each major platform shapes what advertisers can learn from public filings. Knowing which platforms are publicly traded, where they incorporate, and which regulators they answer to is the foundation of platform-policy intelligence.

Deepfake Political Ads 2026 — Platform-by-Platform Detection, Disclosure & Advertiser Liability

Deepfake political ads 2026: where seven platform policies diverge, when FCC and FEC rules apply, and how advertiser liability shifts when synthetic likenesses appear in paid placements.

May 2026 Enforcement Digest: Who Got Banned, Fined or Paused Across 8 Platforms

May 2026 was the heaviest enforcement month of the year so far. Every advertiser-facing action across the eight major platforms — bans, fines, pauses, and the patterns brands need to read before June.

Synthetic Media Enforcement Index Q1 2026 — DSA Transparency Database Findings