Transparency Reports

Last updated: 18.02.2026 · Version: 1.0

This page describes the scope, report types, metrics, and methodology of the Transparency Reports to be published periodically within WIN’s (“Platform”) approach to transparency and accountability.

Transparency reports are prepared both to strengthen user safety and to help ensure that processes relating to content moderation and requests from public authorities are managed in a measurable and auditable manner.

Related documents

Transparency reports are operated together with the following WIN documents as a single coherent “safety and compliance” set:

Report status (current)

Section 1: Purpose and principles

WIN publishes transparency reports for the following purposes:

• Strengthening user trust: Making content moderation and safety processes visible.
• Increasing accountability: Making decisions and processes measurable.
• Detecting risks early: Operational improvement through category-based trends (fraud, harassment, suspected minor safety issues, etc.).
• Supporting regulatory compliance: Transparency obligations under the DSA (Regulation (EU) 2022/2065) in the EU in particular, and process disciplines in the 5651/KVKK context in Türkiye.

The following principles apply when preparing reports:

• Accuracy and consistency: Definitions, counting rules, and periods are stated clearly; comparability is preserved.
• Data minimization and privacy: Personal data is not published; if small numbers would undermine anonymity, suppression/rounding is applied.
• Proportionality: The balance between “how much to report” and security/privacy is observed.
• Accessibility: Reports are published in formats that are as understandable and accessible as possible (see Accessibility Statement).

Section 2: Scope and definitions

Transparency reports generally cover the following event/process classes:

• User report / complaint: Reports made on suspicion of a Community Guidelines violation.
• Illegal content notice: Applications concerning content alleged to be “illegal” (of particular importance in the EU/DSA context).
• Moderation decision: Actions such as content removal, visibility reduction, feature restriction, suspension/closure.
• Appeal/complaint: Internal appeals against moderation decisions (in the EU, DSA Art. 20 and related processes).
• Public authority request: Information/data requests from competent authorities such as courts, prosecution, law enforcement, etc. (see Law Enforcement Guide).
• Orders: In the EU under the DSA, tools such as “orders to provide information” and “orders to act against illegal content”.

Section 3: Publication schedule, format, and access

• Annual: Publication at least once per year is planned (core transparency report package).
• Interim updates (optional): Interim notices may be issued in the event of critical security incidents or legal requirements.

At this stage reports are prepared through manual processes and published as web pages via the Platform’s corporate website and this documentation portal.

3.3 Report packages (map)

Section 4: Report Package A — Community safety and moderation

4.1 Scope

This package covers reports made under the Community Guidelines and moderation actions applied by WIN:

• Profile content (text, photo),
• Messaging content (where applicable and to the extent feasible),
• Behaviour violations (harassment, threats, fraud, etc.),
• Off-platform conduct allegations (limited by an appropriate evidence standard and operational framework).

4.2 Core metric set

4.3 Action types

In this report, actions may be classified at least at the following levels:

• Content actions: removal / visibility reduction / warning label.
• Account actions: warning / feature restriction / temporary suspension / permanent closure.
• Safety actions: automatic protective measures for the reporting party (e.g. mutual invisibility after a report) (depending on product design).

4.4 Appeals (general)

Appeals against moderation decisions are an important indicator of report quality:

• total number of appeals,
• appeal grant/denial rate,
• appeal review time,
• restrictions lifted / decisions reversed after appeal

are reported under a separate subheading.

If EU target: Under DSA Art. 20, at least six months’ internal appeal access and human oversight are applied (see DSA Compliance Page). Outside the EU, the contractual framework: Terms of Use.

Section 5: Report Package B — Illegal content (Notice & Action) and DSA transparency (if EU target)

5.1 Notice & Action metrics (Art. 16)

The report includes at least the following (subject to applicability and exceptions):

• Number of notices under Art. 16 (by category),
• Rate of completion of “minimum elements” for notices (grounds + location/ID + good-faith statement, etc.),
• Rate and time of sending receipt acknowledgements,
• Average/median decision times,
• Use of automation and its relationship to human review.

5.2 Statement of Reasons (Art. 17)

• Number of statements of reasons by restriction type,
• Geographic scope of decisions (within the EU / by country),
• Transparency regarding automation where used,
• Summary of transmission of decisions to the Commission database under Art. 24(5) where applicable in the EU.

5.3 Internal appeal and redress (Art. 20–21)

This sub-report includes at least:

• Appeal counts under Art. 20, grant/denial rates,
• Review times and outcomes,
• Number of applications proceeding to out-of-court dispute resolution under Art. 21 and outcome types.

5.4 Trusted flagger and misuse (Art. 22–23)

• Trusted flagger notice counts and prioritization metrics,
• Counts of suspensions/restrictions relating to repetition of manifestly unfounded notices,
• Counts of sanctions applied to accounts that frequently provide manifestly illegal content.

If there is an EU target, the average monthly active recipients within the EU over the past six months (under Art. 24(2) DSA) will be published; because the application has not yet been released, this figure is for now treated as “0” (zero) or “not applicable”.

Section 6: Report Package C — Government/law enforcement requests and official orders

This package uses the same taxonomy as the Law Enforcement Guide.

6.1 Request types

• Information/document requests (account/profile, technical logs, content data, payment/subscription).
• Traffic data / log requests (in Türkiye, in the 5651 context).
• Preservation / legal hold requests.
• Emergency requests (risk to life / imminent harm).
• Content/account orders:
  • Local procedures in Türkiye,
  • Orders under the DSA in the EU (information provision / action against illegal content).

6.2 Core metric set (recommended)

6.3 User notification (general principle)

As a rule, user notification is aimed at to the extent permitted by law. However, notification may be deferred or omitted if it would jeopardize an investigation, is prohibited by law, or creates a security risk (see Law Enforcement Guide Section 9).

Section 7: Report Package D — Intellectual property notices

In line with the Intellectual Property Policy, this package may include metrics such as:

• numbers of incoming notices (copyright/trademark/other),
• number of items of content removed,
• numbers of counter-notices/appeals,
• repeat infringer enforcement indicators.

As a global reference framework, standards such as DMCA (US) and the DSA notice & action approach in the EU may be considered (see legislative index: DMCA/DSA).

Section 8: Report Package E — Product security

Consistent with the Vulnerability Disclosure Policy, this package may include metrics such as:

• total vulnerability reports,
• severity distribution (critical/high/medium/low),
• remediation times (average/median),
• number of out-of-scope reports (at a high level).

Section 9: Methodology (counting rules and data sources)

9.1 Counting unit: “case”, “content”, “account”

Where possible, reports distinguish three units:

• Case: A report, request, or review file.
• Content: A single content item such as profile photo, profile text, message, media.
• Account: A user account (unique).

This distinction is critical for correct interpretation in situations such as “the same account being reported repeatedly”.

9.2 Period, time zone, and geography

• Each report clearly states the start/end dates it covers and the time zone.
• If geographic breakdown is used, the country/region definition and source (e.g. user declaration, store region, IP-based estimate) are stated.

9.3 Category taxonomy and mapping

• The category set (fraud, harassment, etc.) is fixed.
• If the in-product category set changes, backward comparability is explained in reports with a “mapping” note.

9.4 What does “action taken” mean?

The report distinguishes “action” at least into the following sub-types:

• content action (removal/demotion),
• account action (suspension/closure/feature restriction),
• process action (requesting additional verification, applying preservation, etc.).

9.5 Data sources (examples)

Reports may collect data from:

• In-app reporting/complaint system,
• Support channels (help center / email),
• Legal/compliance channels (official request email/KEP, etc.),
• Moderation decision records,
• Vulnerability reporting channel.

Section 10: Privacy and security safeguards

The following technical/operational measures may be applied during reporting:

• suppression for small numbers or merging categories,
• aggregation/rounding (where necessary),
• reporting only at “outcome type” level,
• access authorization and audit trail.

Section 11: User submissions and feedback

Feedback on transparency reports may be submitted through the appropriate channels:

• For community safety/reporting processes: in-app Report / Notify flow or, for written applications, the Content Violation Form or Complaint and Report Form (see Community Guidelines).
• For public authority requests: official channels under the Law Enforcement Guide.
• For EU/DSA applications: DSA Compliance Page.

Section 12: Changes and version management

WIN may update this page in line with legislation, product, and security needs. The “last updated” information and access to prior versions are published in the in-app “Legal” area and at whoisnextapp.com (opens in a new tab).