What does it mean that 93.6% of moderation is automated, and why does it matter for advertisers?

The figure that 93.6% of platform content-moderation decisions in May 2026 were made by automated means — drawn from the EU DSA Transparency Database — means that the overwhelming majority of ad disapprovals, content removals, and account actions are now decided by algorithms rather than human reviewers, and for advertisers this matters because automated systems generate false positives at scale, removing legitimate content and suspending compliant accounts for reasons that often only a human can correctly assess. At human-review volumes, a wrongful removal was an exception; at 93.6% automation, wrongful removals are a statistical certainty across any large advertising operation. The practical consequences are several. First, speed and opacity: automated decisions are made instantly and often with limited explanation, so an advertiser may discover a disapproval or suspension without an obvious reason. Second, pattern errors: automated systems misread context, flag legitimate use of sensitive keywords, mistake compliant claims for prohibited ones, and apply policy inconsistently across similar creatives. Third, compounding harm: a single automated flag can escalate into account-level restrictions, so an unaddressed disapproval can grow into a suspension. Fourth, the appeal burden shifts to the advertiser: because the platform's first decision is automated, the advertiser must actively invoke the redress mechanisms the law provides rather than wait for a human to notice the error. This is precisely why the EU's Digital Services Act redress framework matters — it gives advertisers a structured path to challenge automated decisions, including the right to a statement of reasons and the right to a complaint that cannot be resolved solely by another algorithm. The defensible posture is to treat automated moderation as a system that will sometimes be wrong, to build a fast appeal workflow, and to prevent flags before they happen by auditing creative pre-publication. To predict and prevent disapprovals use the Meta Rejection Predictor , and for the legal framework see the EU DSA compliance guide . It helps to think in terms of base rates. Even a highly accurate automated classifier — say, one that is wrong only a fraction of a percent of the time — applied to tens of millions of decisions will produce a large absolute number of errors, and a single advertiser running thousands of creatives across markets will encounter several of them. That is not a sign of a uniquely bad platform; it is the arithmetic of automation at scale. The errors also cluster in predictable places: legitimate uses of sensitive vocabulary (a financial-education ad caught by gambling filters, a dermatology ad caught by adult-content filters), regulated-but-compliant products, and creatives that resemble a prohibited pattern without violating it. Because these errors are systematic rather than random, an advertiser can anticipate them and build both prevention and a fast appeal lane in advance, rather than treating each disapproval as a surprise. The teams that suffer least are those that have a named owner for appeals, templates ready, and the evidence to support reinstatement on file before they need it. The organizing principle is that automation at this scale makes wrongful removals routine, and the advertiser must own the appeal.

What is a statement of reasons under DSA Article 17, and what must it contain?

A statement of reasons under Article 17 of the Digital Services Act is the formal explanation a platform must provide to a user when it restricts, removes, demotes, disables, or otherwise acts on their content or account, and it must contain enough detail for the user to understand the decision and pursue redress — which makes it the essential first document in any advertiser's appeal. Article 17 requires that the statement of reasons be clear and specific, and it must include several elements: whether the action affects the visibility, availability, or monetisation of content or the account; the facts and circumstances relied on in making the decision; whether automated means were used in detecting or making the decision, and if so the fact of that automation; the specific legal ground or the contractual or policy provision relied on and an explanation of why the content is considered to violate it; and clear, user-friendly information about the redress options available — the internal complaint-handling system, out-of-court dispute settlement, and judicial redress. The requirement to disclose whether automated means were used is particularly important for advertisers, because if a decision was fully automated, that is itself a basis to demand human review through the internal complaint process. The statement of reasons is also submitted by the platform to the DSA Transparency Database, which is how the public dataset of moderation decisions is built. For an advertiser, the statement of reasons should be preserved immediately upon receiving a disapproval, removal, or suspension, because it identifies the exact policy provision invoked, the facts relied on, and the automation status — the three things an effective appeal must address. An appeal that does not engage with the specific ground cited in the statement of reasons is far weaker than one that rebuts it directly with evidence. To understand how the statement of reasons fits the wider framework see the EU DSA compliance guide , and to track how platforms apply policy see the Policy Change Tracker . In practice, advertisers should extract three things from the statement of reasons and act on each. First, the exact provision cited: the appeal must engage that specific clause, because a complaint that argues against a rule the platform did not invoke will fail regardless of its merits. Second, the facts the platform relied on: these reveal what the automated system actually saw — a particular phrase, an image element, a landing-page attribute — and therefore what the rebuttal must address with evidence. Third, the automation disclosure: if the statement confirms the decision was made by automated means, that is a direct lever, because it supports a demand for the human-supervised review the law guarantees. A practical caution is that statements of reasons are sometimes thin or templated, citing a broad policy area without pinpointing the trigger; where that happens, the advertiser should still preserve the statement and, in the complaint, ask the platform to identify the specific basis while pre-empting the most likely triggers with evidence. Because the statement is also filed to the public transparency database, it is part of an auditable record, which reinforces why a precise, well-documented appeal carries weight. The organizing principle is that the statement of reasons is the appeal's foundation: preserve it, read the cited ground, and rebut it directly.

How does the Article 20 internal complaint system work, and why can't it be fully automated?

The internal complaint-handling system required by Article 20 of the Digital Services Act is a free, accessible, electronic mechanism that platforms must operate so users can challenge moderation decisions, and a central protection is that decisions on these complaints cannot be taken solely on the basis of automated means — appropriately qualified staff must supervise them — which is what gives advertisers a genuine route to human review of an algorithmic error. The mechanics are defined. The system must be available for at least six months from the moment the user is informed of the original decision, so an advertiser has a meaningful window to lodge a complaint. It must be easy to access and user-friendly, and it must enable the submission of sufficiently precise and adequately substantiated complaints — meaning the advertiser should provide specific evidence and argument, not a bare assertion of error. The platform must handle complaints in a timely, non-discriminatory, diligent, and non-arbitrary manner. And the decisive safeguard: where the original decision was made by automated means, the complaint review must not be left to automation alone; qualified human staff must supervise the outcome. If the complaint succeeds, the platform must reverse its decision — reinstate the content, restore the account, or lift the restriction — without undue delay. For advertisers this is the workhorse remedy, because most wrongful disapprovals and suspensions are automated false positives that a human reviewer, presented with clear evidence, can correct. The effectiveness of the complaint depends on its quality: an advertiser should reference the specific ground cited in the statement of reasons, explain precisely why it does not apply, attach evidence such as substantiation for a claim or context for a flagged keyword, and submit within the six-month window. A vague or generic complaint is far more likely to be rejected. To strengthen a complaint by demonstrating creative compliance use the AI Compliance Audit , and for the framework see the EU DSA compliance guide . A few tactical points materially raise the success rate of an Article 20 complaint. Lodge early within the six-month window rather than late, both because automated flags can escalate into account-level restrictions while a complaint sits unfiled and because a prompt, well-evidenced complaint signals seriousness. Make the complaint self-contained: a reviewer should be able to grant reinstatement from the complaint alone, without hunting for context, so attach the substantiation, licence, or contextual explanation directly. Frame the rebuttal around the cited provision and, where the original decision was automated, explicitly request human review on that basis. Keep the tone factual and specific rather than aggrieved; the reviewer is assessing whether the content actually violates the cited rule, not adjudicating a grievance. And preserve the full exchange, because if the internal complaint fails, that record becomes the basis for out-of-court escalation. The human-supervision guarantee is meaningful precisely because it gives a qualified person the chance to see what the classifier missed — but only a complaint that hands that person the evidence makes the guarantee pay off. The organizing principle is that Article 20 guarantees human-supervised review of automated decisions, and a precise, evidenced complaint is the way to win it.

What is out-of-court dispute settlement under Article 21, and when should an advertiser use it?

Out-of-court dispute settlement under Article 21 of the Digital Services Act is an escalation route that lets a user who is dissatisfied with a platform's internal complaint decision take the dispute to an independent, certified dispute settlement body, and an advertiser should use it when the internal Article 20 complaint has failed or stalled and the decision is material enough to justify the effort. The mechanism works as follows. Member-state Digital Services Coordinators certify out-of-court dispute settlement bodies that meet criteria for independence, expertise, accessibility, and impartiality. A user can select any certified body to resolve a dispute about a moderation decision, including disputes that were already submitted through the platform's internal complaint system. Platforms are required to engage with the certified body in good faith, and while the body's decision is not binding in the way a court judgment is, the process is designed to be faster and cheaper than litigation, and platforms bear the fees if the decision favours the user (with the user bearing only limited costs). Importantly, Article 21 does not deprive users of the right to go to court; it is an additional, lower-friction option, not a replacement for judicial redress. For advertisers, the practical decision of whether to use out-of-court settlement turns on materiality and posture: if a wrongful suspension is causing significant ongoing harm and the internal complaint has been rejected on grounds the advertiser believes are wrong, escalation to a certified body is a proportionate next step that signals seriousness without the cost of litigation. If the matter is minor, re-submitting a stronger internal complaint or adjusting the creative may be more efficient. Advertisers should also keep records throughout — the statement of reasons, the internal complaint and its outcome, and all correspondence — because the out-of-court body will assess the dispute on that record. For the framework and the broader set of user rights see the EU DSA compliance guide and to monitor platform policy that underlies disputes see the Policy Change Tracker . A practical point that changes the cost calculus is the fee structure: because platforms generally bear the dispute-settlement body's fees where the decision goes against them, the financial barrier to escalation is low for the advertiser, which makes Article 21 a genuinely usable remedy rather than a theoretical one. That said, the body's decision does not carry the binding force of a court judgment, so its value lies in providing an independent, expert, and relatively fast assessment that platforms must engage with in good faith — useful both to resolve the specific dispute and to create a record. The decision to escalate should weigh materiality against effort: a suspension that is halting a live, revenue-generating campaign justifies escalation, whereas a single disapproved creative with an easy compliant alternative usually does not. Advertisers operating at scale benefit from a standing playbook that defines, in advance, the threshold at which a failed internal complaint is escalated, so the choice is governed by policy rather than made ad hoc under deadline pressure. Throughout, the complete record — statement of reasons, complaint, outcome, correspondence — is what the body assesses, so preserving it from the first notice is essential. The organizing principle is that Article 21 is the proportionate escalation when the internal complaint fails on a material decision, sitting between internal appeal and court.

What is the step-by-step playbook when an ad or account is wrongly removed?

When an ad or account is wrongly removed, the step-by-step playbook is to preserve the statement of reasons, diagnose the cited ground, lodge a precise and evidenced internal complaint within the window, escalate to out-of-court settlement or court if it fails, and prevent recurrence — and following the sequence in order materially improves the odds of a fast reinstatement. Step one is preservation: immediately capture the statement of reasons, the disapproval or suspension notice, screenshots of the creative or content, and the timestamp, because these establish what was decided and on what basis. Step two is diagnosis: read the statement of reasons to identify the exact policy provision invoked and whether the decision was automated; an automated decision is itself an argument for human review. Step three is the internal complaint under Article 20: submit a precise, substantiated complaint that references the specific ground, explains why it does not apply, and attaches evidence — for example, substantiation for a health or financial claim, licensing documentation for a regulated product, or context showing a flagged keyword is benign. Submit within the six-month window and keep the submission. Step four is escalation if the complaint fails: for material decisions, use Article 21 out-of-court dispute settlement through a certified body, or pursue judicial redress; for minor matters, revise and resubmit. Step five is prevention: once reinstated, fix the underlying trigger so the automated system does not flag the same creative again — adjust the copy, add disclosures, or restructure the claim. Throughout, speed matters because automated flags can escalate into account-level restrictions, so the complaint should be lodged promptly rather than after the disruption compounds. Documentation matters because every stage builds on the record. To prevent the disapproval that starts the cycle use the Meta Rejection Predictor , to check copy against risky language use the Keyword Risk Checker , and for the legal framework see the EU DSA compliance guide . Two operational habits separate teams that recover quickly from those that lose campaign time. The first is a named owner: appeals fail or stall most often because no one is clearly responsible for lodging them, so designating an owner — with authority to gather evidence and submit within hours, not days — converts the process from reactive scramble to routine. The second is pre-built assets: maintain templates for the internal complaint, a library of standing evidence (claim substantiation, licences, certification proofs), and a checklist of the most common false-positive triggers for your verticals, so that when a disapproval lands the rebuttal can be assembled and filed immediately. Speed compounds because automated flags can escalate from a single creative disapproval to an account-level restriction if left unaddressed, so the cost of a slow appeal is not just the one ad but the risk to the whole account. Finally, close the loop on prevention every time: a reinstatement that is not followed by a fix to the underlying trigger simply queues the same false positive again. The organizing principle is a disciplined sequence: preserve, diagnose, complain precisely, escalate if needed, and prevent recurrence.

How can advertisers prevent automated flags, and do these rights apply outside the EU?

Advertisers can substantially reduce automated flags by auditing creative before publication, aligning copy with platform policy, substantiating claims, and disclosing material terms — and while the DSA redress rights are an EU framework, the prevention practices are universal and the regulatory direction is global, so a prevention-first posture pays off in every market. On prevention, the most effective lever is pre-publication review, because the cheapest disapproval is the one that never happens. Audit each creative against the platform's prohibited and restricted content rules, screen ad copy for language that automated systems associate with prohibited categories — exaggerated health or financial claims, prohibited superlatives, sensitive-category terms — and ensure any required disclosures are clearly and immediately visible rather than buried. For regulated products, confirm certification and licensing are in place and referenced, because automated systems flag regulated-category ads aggressively. Maintain consistent account health, since a pattern of violations raises the automated scrutiny applied to future ads and can trigger account-level restrictions. On jurisdiction, the specific redress mechanisms — Article 17 statements of reasons, Article 20 internal complaints, Article 21 out-of-court settlement — are obligations under the EU's Digital Services Act and apply to services in the EU and to users in the EU. However, the broader direction is global: Australia's Online Safety Act, the UK's Online Safety Act, and other regimes are building parallel transparency and redress expectations, and major platforms often apply similar appeal mechanisms across markets as a matter of operational consistency. An advertiser operating internationally should therefore assume that documented, evidence-based appeals are the norm everywhere, even where the precise statutory rights differ, and should treat prevention as jurisdiction-agnostic. The asymmetry favours prevention: the cost of auditing creative is trivial against the cost of a suspension during a campaign flight. To audit creative use the AI Compliance Audit and to predict platform-specific rejections use the Meta Rejection Predictor . The economics make the case decisively. Pre-publication review costs minutes per creative and catches the systematic false positives — sensitive-keyword misreads, missing disclosures, uncertified regulated products — before they ever reach an automated reviewer. A disapproval, by contrast, costs the appeal effort plus the campaign time lost while the ad is down, and a compounded account-level restriction can cost a flight entirely. Because automated enforcement scales and is unforgiving, the expected value of prevention is overwhelmingly positive for any advertiser running more than a handful of creatives. The same logic argues for protecting account health as an asset: a clean violation history reduces the automated scrutiny applied to future ads, while a pattern of strikes raises it and can trigger restrictions, so disciplined compliance compounds in the advertiser's favour over time. On jurisdiction, an advertiser should not wait for a market to enact DSA-style rights before adopting evidence-based appeals; the major platforms increasingly apply similar mechanisms across markets, and parallel transparency regimes are emerging, so the documented, prevention-first posture is the right default everywhere. The organizing principle is prevention-first everywhere, with formal DSA rights as the EU backstop and parallel expectations emerging globally.

Automated Moderation 2026: DSA Ad Appeal Rights Guide

Quick Answer

In May 2026, 93.6% of platform content-moderation decisions reported to the EU DSA Transparency Database were made by automated means. At that scale, wrongful ad disapprovals, content removals, and account suspensions are not edge cases — they are an operating reality for advertisers. The EU's Digital Services Act gives recipients of these decisions, including business users and advertisers, concrete redress rights. Article 17 requires platforms to provide a clear statement of reasons for most moderation actions, including whether automated means were used. Article 20 requires platforms to operate a free, accessible internal complaint-handling system, available for at least six months after a decision, and — critically — decisions on those complaints cannot be made solely by automated means; appropriately qualified staff must supervise them. Article 21 lets users escalate to certified out-of-court dispute settlement bodies if the internal complaint fails. Together these create a structured appeal path: read the statement of reasons, lodge an internal complaint with specific evidence, and escalate to out-of-court settlement or judicial redress if needed. For advertisers the practical workflow is to preserve the statement of reasons, document why the decision was wrong, submit a precise complaint within the window, and prevent recurrence by auditing creative before publication. Prevent disapprovals before they happen with the <a href="/tools/meta-rejection-predictor">Meta Rejection Predictor</a> and understand the framework via the <a href="/knowledge/regional-laws/european-union-dsa-compliance">EU DSA compliance guide</a>.

93.6% of Moderation Is Automated: How Advertisers Appeal Wrongful Removals Under the DSA

When Automated Moderation Removes the Wrong Thing

The good news is that the EU's Digital Services Act gives recipients of these decisions — including advertisers and business users — concrete redress rights. Article 17 guarantees a statement of reasons. Article 20 guarantees an internal complaint that cannot be resolved solely by another algorithm. Article 21 provides out-of-court escalation. This guide turns those rights into a practical appeal workflow.

"Providers shall ensure that decisions on complaints are not taken solely on the basis of automated means.
— Digital Services Act, Article 20, internal complaint-handling system"

We cover what the 93.6% number means, the three DSA redress mechanisms, a step-by-step playbook for wrongful removals, and how to prevent automated flags before they happen. To prevent disapprovals use the Meta Rejection Predictor and for the framework see the EU DSA compliance guide.

93.6% Automated: What the Number Means

The automation share is not an abstraction. It changes the probability, speed, and character of the moderation an advertiser faces.

Four Consequences of Automation at Scale

Wrongful removals become routine: At human-review volumes a false positive was an exception; at 93.6% automation it is a statistical certainty across any large operation.
Speed and opacity: Decisions are instant and often thinly explained, so a disapproval or suspension can appear without an obvious reason.
Pattern errors: Automated systems misread context, flag benign sensitive keywords, and apply policy inconsistently across similar creatives.
Compounding harm: A single automated flag can escalate into account-level restrictions, so an unaddressed disapproval can grow into a suspension.

The burden of correction shifts to the advertiser. Because the first decision is automated, you must actively invoke the redress mechanisms the law provides rather than wait for a human to notice. The figures come from the EU DSA Transparency Database; the full breakdown is on the Platform Enforcement Index for May 2026.

Article 17: Your Right to a Statement of Reasons

The statement of reasons is the formal explanation a platform must give when it acts on your content or account. It is the foundation of any appeal.

What It Must Contain

Element	What the platform must disclose
Scope of the action	Whether visibility, availability, or monetisation of content or account is affected
Facts relied on	The facts and circumstances behind the decision
Automation status	Whether automated means were used to detect or decide
Legal / policy ground	The specific provision invoked and why the content violates it
Redress options	Internal complaint, out-of-court settlement, and judicial redress

The automation disclosure matters most: if a decision was fully automated, that is itself a basis to demand human review. Preserve the statement of reasons immediately — it identifies the exact provision invoked, the facts relied on, and the automation status, which are the three things an effective appeal must address. An appeal that does not rebut the cited ground directly is far weaker. For the framework see the EU DSA compliance guide.

Article 20: The Complaint That Can't Be Solely Automated

Article 20 is the workhorse remedy. It requires platforms to operate a free, accessible internal complaint-handling system — with a decisive human safeguard.

How It Works

At least six months to lodge: The system must be available for at least six months from when you were informed of the decision.
Precise and substantiated: Complaints must be specific and evidenced — a bare assertion of error is weak.
Timely and non-arbitrary handling: Platforms must handle complaints diligently and consistently.
Human supervision: Where the original decision was automated, the complaint review cannot be left to automation alone — qualified staff must supervise it.
Reversal without undue delay: If the complaint succeeds, the platform must reinstate the content or account.

This is the route that corrects most automated false positives, because a human reviewer presented with clear evidence can see what the algorithm missed. The effectiveness depends on quality: reference the specific ground from the statement of reasons, explain why it does not apply, attach evidence such as claim substantiation or licensing, and submit within the window. To demonstrate creative compliance in a complaint use the AI Compliance Audit.

Article 21: Out-of-Court Dispute Settlement

If the internal complaint fails, Article 21 provides an escalation route short of litigation.

The Escalation Path

Certified bodies: Member-state Digital Services Coordinators certify independent dispute settlement bodies meeting criteria for independence, expertise, and impartiality.
User choice: You can select any certified body to resolve a dispute, including one already through internal complaint.
Good-faith engagement: Platforms must engage with the body; if the decision favours the user, the platform generally bears the fees.
Court remains open: Article 21 is additional, not a replacement for judicial redress.

Use it when the internal complaint has failed on a material decision causing ongoing harm — it signals seriousness without litigation cost. For minor matters, a stronger re-submission or a creative fix is more efficient. Keep the full record — statement of reasons, internal complaint and outcome, correspondence — because the body assesses the dispute on that record. To monitor the policy underlying disputes see the Policy Change Tracker.

An Advertiser's Playbook for Wrongful Removals

Following the sequence in order materially improves the odds of a fast reinstatement.

The Five Steps

1. Preserve: Capture the statement of reasons, the disapproval or suspension notice, creative screenshots, and timestamps.
2. Diagnose: Identify the exact provision invoked and whether the decision was automated — automation is itself an argument for human review.
3. Complain precisely (Article 20): Reference the ground, explain why it does not apply, attach evidence (substantiation, licensing, benign-keyword context), and submit within six months.
4. Escalate if it fails: Use Article 21 out-of-court settlement or court for material decisions; revise and resubmit for minor ones.
5. Prevent recurrence: Fix the underlying trigger so the automated system does not re-flag the same creative.

Speed matters — automated flags escalate into account-level restrictions, so lodge promptly. To prevent the disapproval that starts the cycle use the Meta Rejection Predictor and to check copy use the Keyword Risk Checker.

Preventing Automated Flags Before They Happen

The cheapest disapproval is the one that never happens. Prevention is jurisdiction-agnostic and pays off in every market.

The Prevention Stack

Pre-publication review: Audit each creative against the platform's prohibited and restricted content rules before launch.
Copy screening: Screen for language automated systems associate with prohibited categories — exaggerated health or financial claims, prohibited superlatives, sensitive-category terms.
Clear disclosures: Make required disclosures clearly and immediately visible, not buried in hover or fine print.
Certification in place: For regulated products, confirm and reference certification and licensing — automated systems flag regulated categories aggressively.
Account health: A pattern of violations raises automated scrutiny on future ads and can trigger account-level restrictions.

The DSA redress rights apply to services and users in the EU, but parallel transparency and redress expectations are emerging in Australia, the UK, and elsewhere, and major platforms often apply similar appeal mechanisms across markets. Assume documented, evidence-based appeals are the norm everywhere. To audit creative use the AI Compliance Audit.

Appeal and Prevention Checklist

[ ] Statement of reasons preserved immediately on any disapproval / removal / suspension
[ ] Cited policy ground and automation status identified
[ ] Internal complaint (Article 20) drafted with specific rebuttal and evidence
[ ] Complaint submitted within the six-month window; submission saved
[ ] Escalation path (Article 21 out-of-court / judicial) assessed for material decisions
[ ] Full record kept: statement of reasons, complaint, outcome, correspondence
[ ] Underlying trigger fixed to prevent re-flagging after reinstatement
[ ] Creative audited pre-publication against prohibited / restricted rules
[ ] Required disclosures clearly and immediately visible
[ ] Regulated-product certification and licensing confirmed and referenced

For multi-jurisdiction stress-testing use the Legal Compliance Scan and for related coverage see the 2026 brand-safety playbook.

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

#Content Moderation#DSA#Account Suspension#Ad Rejection#Appeals#Automation#Statement of Reasons#Advertisers#Platform Liability#Compliance Guide 2026

Share This Report

Tweet Share

Synthetic Media Enforcement Index Q1 2026 — DSA Transparency Database Findings

Q1 2026 DSA Transparency Database snapshot — 299 million enforcement actions across eight major platforms, with the demoted-content layer, automation rates, and EU30 geographic spread broken out.

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.

Cannabis & CBD Social Media Ad Compliance Guide 2026 — Platform-by-Platform Rules, State Laws & Safe Advertising Strategies

A comprehensive 2026 compliance guide to cannabis, CBD, and hemp advertising on every major social media platform. Covers what's allowed, what's banned, topical vs ingestible rules, US state-by-state legal impact, and actionable strategies to run compliant campaigns without getting flagged.

93.6% of Moderation Is Automated: How Advertisers Appeal Wrongful Removals Under the DSA