The answer to this question is short and clear: Internet Law No. 5651 imposes no obligation on a hosting provider to vet comments in advance (pre-moderation). The Law says precisely the opposite. That does not mean "do nothing", however; instead of prior vetting, the Law ties liability to a mechanism of removal upon being put on notice. Below we separate which step is a legal obligation and which is merely good operating practice.
What the Law says: Article 5(1) and 5(2) of Law No. 5651
Law No. 5651, Article 5 — first and second paragraphs
(verbatim; unofficial translation)
ARTICLE 5- (1) The hosting provider is under no obligation to monitor
the content it hosts or to investigate whether unlawful activity is
involved.
(2) (Amended: 6/2/2014-6518/Art. 88) The hosting provider is obliged
to remove from publication unlawful content it hosts where it is put
on notice in accordance with Articles 8 and 9 of this Law.The first paragraph rejects two activities at once: monitoring and investigating whether unlawful activity is involved. Pre-moderation is precisely the sum of those two activities. Not only does the Law fail to impose it as an obligation, it expressly states that there is no such obligation. The same provision is repeated word for word in Article 7(2) of Regulation No. 11746.
- You are NOT required to read comments before they are published — Law No. 5651, Art. 5(1).
- You are NOT required to scan comments regularly, to filter them, or to investigate unlawfulness — Law No. 5651, Art. 5(1).
- You ARE OBLIGED to remove unlawful content from publication once you are put on notice — Law No. 5651, Art. 5(2).
- You ARE OBLIGED to retain traffic data — Law No. 5651, Art. 5(3) (for not less than one year and not more than two years).
Why this two-part structure was built, and why the site's own reporting and a reader's comment are subject to different regimes, is set out in detail together with the text of the provisions in our article on whether a website is liable for reader comments.
Pre-moderation and post-moderation: the difference in regime
The regime established by Law No. 5651 is one that operates after the fact. The event that gives rise to the obligation is not the writing of the comment but your being put on notice. Pre-moderation is not part of that regime; it is the business's own choice.
| Issue | Pre-moderation | Post-moderation (the regime of the Law) |
|---|---|---|
| Statutory basis | None — no such obligation is laid down in Law No. 5651 | Law No. 5651, Art. 5(2) — removal upon being put on notice |
| Moment the obligation arises | — | Arrival of a notification in due form |
| Timing | Before publication | After the notification |
| The Law's stance | Expressly placed outside the scope of obligation by Art. 5(1) | Laid down as an obligation |
| Operating cost | High — every comment is read by a human | Low — action is taken when a notification arrives |
Does pre-moderation move you closer to the position of a content provider?
This is the most debated question in the sector. The honest answer: Law No. 5651 contains no express provision on the point. The Law contains nothing at all to the effect that a hosting provider carrying out pre-moderation changes its capacity or loses the protection in Article 5(1). Because there is no such provision, we do not state a definite conclusion here.
The only test to be found in the Law is the definition of the capacities. Article 2(1)(f) defines the content provider as "natural or legal persons who produce, modify and provide any kind of information or data offered to users over the internet"; Article 2(1)(m) defines the hosting provider as persons who "provide or operate the systems hosting services and content". The legal debate therefore turns on whether a site that carries out pre-moderation has in fact moved into the position of a party that "modifies and provides". We have examined every dimension of this distinction between capacities in our article on which capacity a news website holds.
What exactly does "being put on notice" mean?
Article 5(2) makes the removal obligation conditional on "being put on notice in accordance with Articles 8 and 9 of this Law". The Law also contains a definition of a "warning method":
Law No. 5651, Article 2(1)(r) (verbatim; unofficial translation)
r) (Added: 6/2/2014-6518/Art. 85) Warning method: the method of
notification to be carried out by persons claiming that their
rights have been infringed by reason of the content of a
publication made on the internet, for the purpose of having the
content removed from publication, addressed first to the content
provider and, where no result is obtained within a reasonable
time, to the hosting provider, through their contact addresses,What to do once a notification has arrived, and in which cases meeting the request and in which cases refusing it produces consequences, is set out step by step in our article on the route to follow when a removal request arrives and in our article examining the consequences of refusing a request.
Mandatory versus recommended: the clear dividing line
| Practice | Status | Legal basis |
|---|---|---|
| Reading comments before publication | NOT a legal obligation | Law No. 5651, Art. 5(1) — there is no monitoring or investigation obligation |
| Scanning comments regularly, setting up filters | NOT a legal obligation — an operational choice | Law No. 5651, Art. 5(1) |
| Publishing comment rules | NOT a legal obligation — an operational recommendation | Not regulated in Law No. 5651 |
| Setting up a notification channel | An indirect obligation | Law No. 5651, Art. 3(1) identifying information + Art. 2(1)(r) warning method |
| Keeping identifying information up to date | A LEGAL OBLIGATION | Law No. 5651, Art. 3(1) — sanction under Art. 3(2): TRY 2,000 – 50,000 |
| Removing content once put on notice | A LEGAL OBLIGATION | Law No. 5651, Art. 5(2) |
| Retaining traffic data (IP, port, time) | A LEGAL OBLIGATION | Law No. 5651, Art. 5(3) + Art. 2(1)(j) |
| Keeping a removal log | Not separately regulated in the Law — a means of proof | — |
The "indirect obligation" row in the table is so described for the following reason: the Law does not tell you to "put up a notification form"; but Article 3(1) requires identifying information to be kept accessible to users and up to date, and the warning method in Article 2(1)(r) says that the notification is to be made "through their contact addresses". On top of that, under Article 6/A(11) the Access Providers Association may notify its decisions to an e-mail address that can be identified from the web pages of the content or hosting provider. In practice, therefore, your contact address is an official notification channel.
Traffic data: what is retained, and for how long?
Under Article 5(3) of Law No. 5651 the hosting provider is obliged to retain the traffic data relating to the services it hosts for a period to be determined by regulation, being not less than one year and not more than two years, and to ensure its accuracy, integrity and confidentiality. Traffic data is defined in Article 2(1)(j): the IP address, port information, start and end time of the service provided, the type of service used, the amount of data transferred and, where applicable, the subscriber identity information relating to the parties.
The notification channel: minimum fields
When designing your notification form you can take as a model the information required by Article 9/A of Law No. 5651 for an application concerning privacy of private life. The second paragraph of that article says that the request is to include "the full address (URL) of the publication causing the infringement of the right, an explanation of the respects in which the right has been infringed, and information proving identity", and that where any of this information is missing the request will not be processed. Making those same three elements the minimum for your own form both guides the applicant and documents your position.
COMMENT NOTIFICATION FORM — MINIMUM FIELDS
[ ] Full internet address (URL) and comment ID
[ ] A copy of the text of the comment complained of
[ ] Explanation of the respect in which the right is infringed
[ ] Name and surname / title of the applicant
[ ] Information proving identity
[ ] Contact address and telephone
[ ] Power of attorney, where an attorney applies
[ ] Number and date of any judicial or administrative decision
[ ] Date and time of the application (recorded automatically by the
system)
FIELDS RECORDED AUTOMATICALLY (on the operator's side)
- Application number
- Date and time the notification arrived (Turkish time)
- Person carrying out the assessment
- Decision: removed / not removed / partially edited
- Date and time of the decision
- Grounds
- Date and content of the reply given to the applicantTwo frequently asked questions
If I close comments altogether, does my liability come to an end?
If you close the comment area, you no longer host third-party content in that area, and the basis for your hosting provider capacity on that leg falls away. But as regards the site's own reporting, your content provider liability (Law No. 5651, Art. 4(1)) and the chain of criminal and civil liability in Articles 11 and 13 of Press Law No. 5187 continue exactly as before. Closing comments does not reduce the liability arising from editorial content.
If I carry out pre-moderation, am I deemed to have been "put on notice"?
The Law has not regulated this. Article 5(2) ties being put on notice to an event occurring "in accordance with Articles 8 and 9"; there is no provision in the Law as to whether the fact that your own moderator has read a comment counts within that scope. You will have to make your decision accepting that uncertainty — we are not presenting you with a definite conclusion that is not in the text of the Law.
Comment policy template
The text below is not a legal obligation; it is an operational tool that clarifies the relationship between you and the reader and documents the site's position in the event of a dispute. Adapt it to your own site.