OLG Frankfurt am Main - 16 W 37/20

From GDPRhub
OLG Frankfurt - 16 W 37/20
Courts logo1.png
Court: OLG Frankfurt (Germany)
Jurisdiction: Germany
Relevant Law: Article 6(1)(f) GDPR
Article 17 GDPR
Article 85(2) GDPR
Decided: 19.11.2020
Published:
Parties:
National Case Number/Name: 16 W 37/20
European Case Law Identifier: ECLI:DE:OLGHE:2020:1119.16W37.20.00
Appeal from: LG Frankfurt (Germany)
2-03 O 167/20
Appeal to: Not appealed
Original Language(s): German
Original Source: Bürgerservice Hessenrecht (in German)
Initial Contributor: Agnieszka Rapcewicz

The Higher Regional Court in Frankfurt am Main (OLG Frankfurt am Main) dismissed an appeal brought by a doctor against an online evaluation portal, holding that it was in the public interest that the doctor's profile be labelled with a warning sign because of the reviews which the doctor had received.

In this case however, the Court stated that the dispute doesn't concern the processing of personal data, but rather an independent statement by the defendant, with the warning of the defendant, concerning possibly manipulated assessments. Therefore the alleged violation of GDPR provisions were unfounded.

English Summary[edit | edit source]

Facts[edit | edit source]

The complainant was a doctor listed on the defendant's evaluation website. In February 2020, the defendant informed the claimant that he had discovered that positive assessments, so-called false assessments, had been published on his profile and asked him to clarify this situation. The sanction for the lack of explanation was supposed to be information for users of the portal in the form of a warning on the doctor's profile that reviews had been purchased. In the correspondence exchanged, the doctor denied that there were false assessments on his profile, but the defendant found it not sufficient. The defendant then posted a warning notice on the applicant's profile. The plaintiff sought a temporary injunction to prohibit the marking of his profile with a warning sign and the display of the text of the warning. The Regional Court dismissed the application. The applicant challenged that decision and filed an immediate appelation. The Regional Court did not remedy the complaint and referred the matter to the Higher Regional Court for a decision.

Dispute[edit | edit source]

Does the public interest invoked by the defendant in connection with the publication of the warning outweigh the interests of the applicant? Has the defendant infringed the provisions of the GermanCivil Code in conjunction with the Article 6(1)(f) GDPR, and could not invoke the press exception for the publication of the warning?

Holding[edit | edit source]

The Higher Regional Court declared the appeal to be unsuccessful. It held that the plaintiff has no right against the defendant to stop the labelling of his profile with a warning sign and the display of the warning sign.

Comment[edit | edit source]

The allegation that the defendant violated the German Civil Code in conjunction with Article 6(1)(f) was only one of the arguments of the plaintiff. The defendant claimed that the contested warning notice is a contribution by the defendant to the formation of users' opinions, which is why the defendant may rely on the media privilege. The applicant considered that, in weighing up the public interest and the interest of the applicant, the defendant, while processing personal data, could not rely on the privileged power to publish a warning. That is the privilige of the press. The plaintiff referred in this respect to decisions of the Cologne Higher Regional Court and the Federal Court of Justice. The plaintiff claimed to be entitled to a right to have his data not processed in the warning notice also arising from Sections 823 (2) and 1004 of the German Civil Code by analogy in conjunction with Sections 823 (2) and 1004 of the German Civil Code. The plaintiff claimed that Article 6(1)(f) GDPR was applicable in the present case, since the defendant was relying (in the applicant's opinion) on the so-called media privilege. In the applicant's opinion there was no public interest in the publication of the warning notice in the present case and the defendant had not put forward any arguments to show that the applicant was responsible.

The Higher Regional Court found that irrespective of whether Article 6 GDPR could be regarded as a protective law within the meaning of Section 823 (2) of the German Civil Code (BGB) or whether a claim for injunctive relief under data protection law falls under Article 17 GDPR, the applicability of Article 6(1)(f), which governs the lawfulness of data processing, is ruled out because there is no data processing in the present case. Pursuant to Article 2(1) GDPR, the Regulation applies to the wholly or partially automated processing of personal data. According to Article 4(2) GDPR, the term "processing" refers to any operation or set of operations carried out with or without the aid of automated procedures in connection with personal data, such as collection, recording, organisation, organisation, filing, storage, adaptation or alteration, reading, retrieval, use, disclosure by transmission, dissemination or any other form of making available, alignment or combination, restriction, deletion or destruction.

The Court held that it is not the deletion or restriction of personal data that is the subject of the dispute, but rather an independent statement by the defendant, with the warning of the defendant, concerning possibly manipulated assessments. Incidentally, in its decision, which concerned the classification of ratings by the internet portal as "recommended" or "not recommended", the Federal Court of Justice did not apply the GDPR, but Sections 824, 823 (1), 1004 (1) of the German Civil Code (BGB) (BGH, ruling of 14 January 2020, VI ZR 496/18). In the absence of applicability of the GDPR, it remains open whether the defendant could invoke the media privilege under Article 85 GDPR.

Further Resources[edit | edit source]

Share blogs or news articles here!

English Machine Translation of the Decision[edit | edit source]

The decision below is a machine translation of the German original. Please refer to the German original for more details.

Warning notice on physician portal lawful in case of suspicion of manipulated evaluation

Orientation kit
A physician rating portal may mark the physician profile with a warning if there is a justified suspicion of "bought" ratings.

Note
The decision is not subject to appeal.

A press release on this decision is available on the website of the OLG (www.olg-frankfurt-justiz.hessen.de)

Proceedings
Forthcoming LG Frankfurt am Main, 9 June 2020, 2-03 O 167/20, Decision
Tenor
The immediate appeal of the plaintiff against the order of the Regional Court of Frankfurt am Main of 9 June 2020, ref. 2-03 O 167/20, is dismissed.

The plaintiff shall bear the costs of the appeal.

The amount in dispute for the appeal is set at € 20,000.

Reasons
A.

The plaintiff (hereinafter: plaintiff) is ...physician in Stadt1 and as such is listed in the evaluation portal of the defendants (hereinafter: defendant), whose contractual partner he was until 16 April 2020. On 17 February 2020, the defendant informed the plaintiff, giving examples of two evaluations, that it had discovered that positive evaluations, so-called fake evaluations, had been published on his profile and requested him to clarify how this could have come about. In the event of non-enlightenment, the defendant threatened to inform the users by means of a warning notice on the profile that there were purchased reviews on his profile (cf. email Bl. 30 f. d.A.). After the subsequent correspondence, in the course of which the plaintiff rejected the accusation that he had bought fake positive ratings and published them on his portal, the defendant published a warning notice on the plaintiff's profile. The notice is not displayed in the results list of a doctor search; the warning notice is only visible to the user when he accesses the plaintiff's profile. In this case, the warning is initially indicated by a small exclamation mark with a red background in the top left-hand corner of the overall grade. When the user moves the mouse over the overall mark, the following text opens:

"We have noticed conspicuous features in individual evaluations on this profile which cause us to doubt their authenticity. We have confronted the profile owner with the facts. As a result, the matter could not be clarified so far. The profile owner denies being responsible for the manipulation.

In order to give users an idea of the credibility of the ratings of a profile, we mark profiles where there are suspicions of manipulation in the form of purchased or unfairly influenced ratings. Whether the manipulations were initiated by the profile owner, we cannot at present make a final judgement, despite having contacted him.

We are constantly developing our procedures to identify manipulated ratings, remove them and take decisive action against those responsible. Nevertheless, it cannot be ruled out that this profile will continue to contain manipulated ratings in the future.

The plaintiff seeks a temporary injunction to prohibit the marking of his profile with a warning sign and the display of the text of the warning.

In the first instance, the plaintiff asserted that the text of the notice was untrue factual assertions. The accusation was made specifically that his profile contained bought or manipulated ratings and that he was responsible for them. The average reader understands the warning in such a way that only the plaintiff can be responsible for the allegedly bought or manipulated ratings. Whether the plaintiff had bought or manipulated ratings on his profile could be verified by means of evidence.

The assertion that the defendant was unable to prove the truth is untrue. Nor could the defendant prove that the applicant's profile contained any false evaluations.

The qualification as an untrue factual claim also follows from the selective provision of information. There is no mention of the fact that in the case of anonymous and brief evaluations, it is not possible for the plaintiff to understand who gave these evaluations. Nor was it stated that the assessments could have been made by criminals who wanted to blackmail the plaintiff by passing the data on to the defendant.

The principles of reporting suspicions have not been complied with. In view of the serious allegations of manipulation of assessments, the highest standards of due diligence in reporting had to be applied.

It was not clear from the note in which period of time conspicuousness had allegedly occurred. The reporting was prejudicial, to which the presentation of the warning notice also contributed. The applicant's comments were not duly taken into account, there is no minimum of evidence and the defendant cannot rely on the defence of legitimate interests.

The Regional Court dismissed the application. In support of its application, it essentially stated that the statements made did not affect the plaintiff's interests, which are protected by fundamental rights. In the context of the weighing of interests to be carried out, the interest of the defendant as operator of the rating portal in the provision of authentic physician ratings outweighs the interest of the defendant.

The challenged statements were factual allegations. The defendant's statement in dispute was admissible in accordance with the principles of suspicion reporting with regard to the circumstances of the present individual case.

There is a minimum amount of evidence to support the suspicion that the plaintiff had purchased valuations or that a purchase could be attributed to him. The defendant discovered, on the basis of e-mail and IP addresses, that evaluators had worked for evaluation providers and that these evaluators had also evaluated the applicant's medical profile. The fact that these users had given purchased reviews was admitted by other doctors who had been rated by these users. In addition, the present case concerns a relatively large number of more than 10 reviews which were purchased following the defendant's investigation.

According to the applicant, the allegations had not been sufficiently addressed either before the court or on judicial advice. He did not disclose what research he had undertaken to clarify the facts of the case, nor did he sufficiently contest his responsibility for the valuations which he appeared to have purchased. In so far as the plaintiff argues that all valuations must be those of blackmailers who had sent him letters in the past, these statements are not sufficient to refute the suspicion formulated in the information in dispute in the long term. The defendant has shown that it has no complaints from users alleging that they are being blackmailed by positive evaluations.

In view of the numerous evaluations that were manipulated according to the defendant's research, it is not clear whether the evaluation of 23 May 2019 is based on a treatment contact, as the plaintiff claims.

It is also in the public interest that the defendant shows the warning notice in the plaintiff's assessment profile. There is a public interest in a functioning and credible evaluation platform.

The defendant's warning notice was also not prejudicial. Rather, it adequately describes the state of the investigations and sufficiently expresses that it is not an established fact but a mere suspicion.

The defendant had sufficiently heard the applicant before publishing the warning. It may be that the defendant did not disclose in detail all the assessments suspected of manipulation. However, it was sufficient in that regard that, for the purposes of the hearing, it referred to two evaluations considered to be exemplary and described in summary form the phenomenon of purchased evaluations and the evidence relating to the applicant's profile.

In the context of the balancing of interests, the Board also took account of the fact that the applicant's profile was a paid profile with the additional options 'platinum' and 'top placement' during the period of the evaluations at issue in the present case.

The plaintiff objects to that by his immediate appeal. He submits that the privileged conditions for reporting suspicions are not applicable in the present case. The defendant is abandoning its role as a neutral mediator of information, since it is expressing its own suspicions and not merely listing the assessments of others. When weighing up the interests, this had to be taken into account at the expense of the defendant. This meant that the defendant could not rely on the privileged conditions of suspicious activity reporting, which privilege the press. The plaintiff refers in this respect to decisions of the Cologne Higher Regional Court and the Federal Court of Justice. The plaintiff is therefore entitled to a right to have his data not processed in the warning notice also arising from Sections 823 (2) and 1004 of the German Civil Code by analogy in conjunction with Sections 823 (2) and 1004 of the German Civil Code. Article 6(1)(f) DSGVO. Article 6(1)(f) DSGVO is applicable in the present case, since the defendant cannot rely on the so-called media privilege.

Apart from the principles of reporting suspicions, there is no justification for such a disparaging denunciation.

At the very least, the defendant had infringed the principles of reporting suspicions, which the plaintiff reiterates in detail.

There was just no sufficient evidence that purchased evaluations were available on the plaintiff's profile and could be attributed to him. The defendant merely submits that doctors have admitted that they bought reviews, but not that the user specifically listed published a falsified review. The fact that doctors admitted that they had bought reviews is not to be equated with the fact that the specific users submitted purchased reviews and published purchased reviews on the applicant's profile. The applicant has provided prima facie evidence that the rating k) originates from a ... patient of the applicant. In the case of the other reviews, the defendant had not proved that these users had actually submitted false reviews.

The plaintiff had sufficiently cleared up the accusations. The Regional Court did not take sufficient account of the fact that the plaintiff had provided prima facie evidence that he had received several letters from blackmailers threatening to send the data of the positive ratings already published to the defendant if he did not pay €500. The truthfulness of the affidavit was not contradicted by the defendant's statement that it had no complaints from users that they were being blackmailed by positive reviews.

The Landgericht is wrong to assume that the plaintiff should have disclosed what research he had undertaken to clarify the facts of the case. The plaintiff submitted that, due to the lack of information in the ratings, he had no further indications as to the patients from whom the ratings originated. Also, because of the time that had elapsed since then, it was no longer possible for the applicant to establish who had written which evaluation. The plaintiff was only able to attribute one evaluation to one patient. The fact that the evaluation of the ... patient was not purchased was significant for the decision and could not be ignored. The plaintiff had affirmed in lieu of an oath that he had never bought or manipulated evaluations and that he had not paid any persons or companies to publish falsified evaluations. In addition, the plaintiff submits a further affidavit (annex, p. 232).

Furthermore, it is to be seen that the plaintiff not only booked the top placement with the defendant, but that with more than 80 evaluations, he was always awarded the grade 1.0. The defendant did not want to claim that all of the plaintiff's evaluations were falsified. It would have been superfluous to buy manipulated evaluations.

There was no public interest in the publication of the warning notice in the present case and the defendant had not put forward any arguments to show that the applicant was responsible.

The defendant's warning notice was also prejudicial. Contrary to the view of the Court of First Instance, the average reader of the warning assumes that only the applicant can be held responsible for the manipulated assessments.

The defendant did not sufficiently hear the applicant before publishing the warning.

When weighing up the interests, the general right of personality and the right in the plaintiff's established and practised business enterprise outweighed the right to express the defendant's opinion. The interest of the defendant in taking action against fake valuations could at best go as far as deleting proven fake valuations. However, the public denunciation of doctors was clearly going too far. Moreover, it is not clear for how long the reference will remain published on the plaintiff's profile. In view of the fact that the assessments were allegedly purchased in 2019, publication in mid-2020 was no longer appropriate. Moreover, the average user thought that even now there would still be manipulated ratings on the profile. In fact, according to the defendant's own statements, it had deleted all conspicuous ratings.

The Regional Court did not remedy the complaint and referred the matter to the Higher Regional Court for a decision.

In a written statement of 8 July 2020, the plaintiff additionally asserts, with reference to a decision of the District Court Stadt2 (ruling of 15 June 2020, ...), that a right to injunctive relief also results from Sections 280 (1), 241 (2) BGB. By publishing the warning notice, the defendant had culpably breached the duties of protection and loyalty arising from the contractual relationship between the parties.

In a further written statement dated 13.7.2020, the plaintiff submitted an affidavit by his practice manager, Ms A, responsible for Bewertungsportal1 , who was also not responsible for allegedly purchased or manipulated ratings.

The defendant defends the contested decision.

The contested warning notice is a contribution by the defendant to the formation of users' opinions, which is why the defendant may rely on the media privilege.

Doctor B from Stadt1 admitted to having bought ratings. On his screenshot from his customer account at GoldStar Marketing, users C and D appeared whose data coincided with the data in the ratings for c and g on the plaintiff's profile.

The claimant's statement concerning an alleged ... patient is unbelievable and the affidavit is incorrect in this respect. Even in the case of the alleged blackmail attempts with positive evaluations, the plaintiff only described vague facts.

The claim for injunction could not be based on the possible conclusion of the reader that the plaintiff himself was responsible for the manipulation. In this respect, there was no hidden statement.

By means of the warning notice, the defendant complies with the Bundeskartellamt's demand to inform consumers.

It is not a permanent warning notice. Since the source of the manipulation has not been uncovered, the defendant must expect that new non-authentic reviews will be posted at any time, especially since purchased reviews are not suddenly put in the profile. As manipulated ratings were deleted, the defendant had to reckon with the fact that the service provider would try to post new ratings. After setting the warning notice, the defendant evaluates the respective profile over a certain period of time.

In addition, the defendant comments on the decision of the LG Stadt2. In addition, it argues that it has been confirmed in the past that only the publication of the warning notice leads to a serious clarification of the facts by the doctors contacted. Placing the warning notice is the only way to ensure that the facts are clarified.

B.

The applicant's immediate appeal is admissible, but is unsuccessful on the merits.

The plaintiff has no right against the defendant to stop the labelling of his profile with a warning sign and the display of the warning sign.

I. Without success, the plaintiff initially asserts a claim for injunction under Sections 280 (1), 241 (2) of the German Civil Code (BGB) with reference to the decision of the Regional Court Stadt2 of 15 June 2020, file no. ..., on the grounds of a culpable breach of the duties of protection and loyalty arising from a contractual relationship.

1 In the case decided by the Regional Court of Stadt2 , the parties to the contract were bound by a contract which had been terminated in the meantime but had not yet been terminated at the time of the last oral hearing. It is unclear here whether a contract between the parties (still) exists. In his application, the plaintiff submitted that he was the defendant's contractual partner until 16 April 2020, so that after that date no culpable breach of an existing contractual relationship could be assumed. However, the defendant stated in a written statement dated 24 July 2020 that a "platinum" contractual relationship existed between the parties to the effect that the plaintiff could upgrade his profile in return for payment of a fee and had the opportunity to write articles in the Expert Advisor, to be displayed more prominently in an advertisement window for specialist areas and to make use of the Profile Service.

2 However, even if a contractual relationship were to be accepted as submitted, the Senate is not able to recognise that there would be ancillary duties of protection and loyalty which would prohibit the defendant from publishing a warning notice, even if, as the Landgericht Stadt2 assumed, it were untrue. It is true that the contractual obligation under Section 241 (2) of the German Civil Code (BGB) also gives rise to duties of consideration and protection, in particular the duty to behave in such a way when settling the obligation that the body, life, property and other legal interests of the other party are not injured. However, the warning notice is not displayed during the settlement of the contractual obligation (see also OLG Munich, judgement of 27.2.2020, 29 U 2584/19, juris, for the deletion of positive ratings in dispute there). The contractual obligations of the defendant relate solely to the individual design of the plaintiff's profile, its presentation and special services. The contractual relationship does not include transactions concerning the ratings, with regard to the treatment of which the defendant makes no distinction between contractual partners and other doctors. In addition, the respective physician who makes use of separate paid services of the defendant is also informed at the beginning of the booking process with the remark that the booked package has no influence on the evaluations. A lack of influence on the evaluations means at the same time that the defendant is not subject to any contractual obligation to show consideration with regard to the evaluations. Without success, the applicant submits that, under the contract, the defendant also owes the provision of the profile, including the ratings given on it, since the further services of the Platinum Premium Package are otherwise not possible. Even if the Platinum Package is based on the - non-payable - provision of the profile, the (ancillary) contractual obligations do not extend to this. In addition, if a duty of protection and loyalty is accepted by booking a package, the doctor could "buy his way out" of the fact that the defendant will make a reference to this if there is a justified suspicion of manipulated ratings. Such a difference in treatment between doctors as customers and non-customers would not be compatible with the defendant's basic role as a provider of information.

Therefore, even with an "untrue" warning notice (according to the Landgericht Stadt2 ), the defendant would not have breached duties of protection and loyalty towards the plaintiff.

II Nor does the plaintiff's claim to a right of disposal follow from Sections 823 (2) and 1004 of the German Civil Code by analogy in conjunction with Article 6, paragraph 1 f) DSGVO.

With regard to this basis for the claim, the plaintiff refers to a decision of the Cologne Higher Regional Court (OLG Cologne, judgment of 14 November 2019, I-15 U 126/19). Irrespective of whether Art. 6 DSGVO can be regarded as a protective law within the meaning of Section 823 (2) of the German Civil Code (BGB) or whether a claim for injunctive relief under data protection law falls under Art. 17 DSGVO, the applicability of Art. 6 (1) f DSGVO, which governs the lawfulness of data processing, is ruled out because there is no data processing.

Pursuant to Article 2(1), the DSGVO applies to the wholly or partially automated processing of personal data. According to Section 4 No. 2 of the DSGVO, the term "processing" refers to any operation or set of operations carried out with or without the aid of automated procedures in connection with personal data, such as collection, recording, organisation, organisation, filing, storage, adaptation or alteration, reading, retrieval, use, disclosure by transmission, dissemination or any other form of making available, alignment or combination, restriction, deletion or destruction.

It is not the deletion or restriction of personal data that is the subject of the dispute, but rather an independent statement by the defendant, with the warning of the defendant, concerning possibly manipulated assessments. Incidentally, in its "yelp" decision, which concerned the classification of ratings by the internet portal as "recommended" or "not recommended", the Federal Court of Justice did not apply the DSGVO, but Sections 824, 823 (1), 1004 (1) of the German Civil Code (BGB) (BGH, ruling of 14 January 2020, VI ZR 496/18).

In the absence of applicability of the DSGVO, it remains open whether the defendant could invoke the media privilege under Art. 85(2) DSGVO.

III Finally, the plaintiff also has no right of injunction against the defendant under Sections 823 (1), 1004 (1) of the German Civil Code (by analogy), Article 1 (1), 2 (1) and 12 of the Basic Law.

1 The warning notice affects the scope of protection of the plaintiff's general right of personality under Article 1 (1) and (2) (1) of the Basic Law and his right to an established and operated business under Article 12 (1) of the Basic Law, which protects the plaintiff's enterprise in its economic activity and in its functioning from unlawful interference by acts of infringement directed against the enterprise and its organisation or against the entrepreneurial freedom of decision (BGH, ruling of 14 January 2020, VI ZR 496/18 marginal no. 35).

The warning notice together with the text is likely to have an impact on the personal and corporate reputation of the plaintiff and may lead to patients who inform themselves about the defendant's platform deciding against the plaintiff as a treating physician.

2. however, the intervention is not unlawful.

a) Without success, the plaintiff claims that the warning notice is an untrue statement of fact. In the first instance, the plaintiff complained about the existence of untrue factual allegations to the effect that there were purchased or manipulated ratings on his profile and that he was responsible for them; the average reader understands the warning notice in such a way that only the plaintiff can be responsible for the allegedly purchased or manipulated ratings. In the complaint, the plaintiff first of all intervened solely on the question of the admissibility of the suspect reporting. On presentation of the decision of the District Court Stadt2 , however, he then again took the view that there was an "untrue warning": The targeted public understood the warning in such a way that the profile owner himself was responsible for the allegedly manipulated evaluations. The defendant presents him as a liar and cheat.

However, the warning notice cannot be understood in this sense.

Its first paragraph explicitly states that the profile owner denies being responsible for the manipulation. In the second paragraph, the warning also states that, despite having been contacted, the defendant has not been able to establish definitively whether the manipulations were initiated by the profile holder. Although it is further stated that the defendant is constantly developing its procedures to identify manipulated ratings and is taking decisive action against "those responsible", it cannot be ruled out that this profile still contains or will contain manipulated ratings. In this respect too, however, only the "profile" and not the profile owner as the person responsible is taken into account.

Nor does the warning notice contain any concealed statement to the effect that the plaintiff had actually initiated the alleged manipulations. A "concealed" statement, which is equivalent to an "open" statement, can only be assumed if the author makes an additional factual statement through the interaction of open statements or suggests it to the reader as an unrefutable conclusion (BGH, judgement of 22.11.2005, VI ZR 204/04). Such an unrefutable conclusion is not present here. The warning explicitly states that the defendant was currently unable to clarify definitively whether the manipulations were initiated by the profile owner. This raises the possibility of third party liability, even if it is not explicitly mentioned.

Finally, the principles of the so-called Stolpe case-law (BVerfG, order of 25 October 2005, 1 BvR 1696/98), which the plaintiff invokes in a written statement of 20 October 2020, do not lead to the understanding alleged by the plaintiff to be unfavourable to him. The reader may well conclude - albeit not unrefusably - that the plaintiff is responsible for manipulation. However, this does not mean that the warning notice itself is ambiguous as to the responsibility of the plaintiff. Rather, it reports a suspicion that may or may not be true.

Ultimately, the plaintiff also assumes that it depends on whether the principles of suspicion reporting have been complied with (cf. p. 358 of the Code of Conduct, although it is not mentioned elsewhere - see below - at the same time, he takes the view that these principles are not applicable).

b) The principles of suspicion reporting have been observed.

aa) The principles of suspicion reporting are applicable.

The plaintiff denies the applicability on the grounds that the defendant, with the warning notice, leaves its position as a neutral information intermediary, so that it cannot fully invoke its freedom of opinion. In this respect, he refers to a decision of the Higher Regional Court (OLG) of Cologne of 14 November 2019, 15 U 126/19 (see already above under II.), in which it was a matter of a doctor's application for deletion of the data published on his profile without his consent and of a failure to publish this data in connection with a different presentation in the case of paying customers. In that case, the Higher Regional Court (OLG) of Cologne stated, in the context of the discussions on the intervention of the media privilege, that the defendant could not "change roles" and mutate from a mere information intermediary with only a reactive duty to verify to a disseminator of his own opinions who could comprehensively invoke freedom of opinion and freedom of the press. The plaintiff also refers to the decision of the Federal Court of Justice "Jameda III" (ruling of 20 February 2018, VI ZR 30/17, marginal no. 19), in which the Federal Court of Justice stated that the defendant could only assert its legal position based on the fundamental right of freedom of opinion and freedom of the media - there: compared with the right to protection of personal data - with less justification if it withdrew in its role as a "neutral" information intermediary in favour of an advertising offer.

However, it is not apparent from the cited decisions that the principles of suspicion reporting could not apply to the defendant. The decision of the Cologne Higher Regional Court deals with the media privilege of the defendant. Even if it were not subject to it, it does not mean that it could not invoke its fundamental rights under Article 5 of the Basic Law or - as the plaintiff puts it - the "conditions of permissible suspicious reportings which privilege the press". Moreover, the facts on which the decisions of the Cologne Higher Regional Court and the Federal Court of Justice were based are not comparable with the facts here. In the present case, it is not a question of the defendant conferring an advantage on paying customers through a particular design of its site. Rather, the defendant's aim is to preserve the functionality of its portal by trying to prevent bought and manipulated applications. The operation of a rating portal, which is covered by the scope of protection of Article 5 (1) sentence 1 of the Basic Law (BGH, ruling of 23.9.2014, VI ZR 358/13 marginal no. 28), fulfils a function which is approved by the legal system and is socially desirable. This also applies to a control and evaluation of user contributions by the operator of a rating portal for the purpose of protecting and supporting its functionality. This is because there is a general risk of untrue, insulting or otherwise inadmissible statements and the misuse of the evaluation portal through the posting of multiple evaluations by one and the same person as well as evaluations without a real background of experience. A rating portal can also contribute to the formation of opinion among users who inform themselves by assessing user contributions themselves - (also) taking the above-mentioned aspects into account or independently of them (BGH, judgement of 14.1.2020, ibid.). By taking measures to counteract the abuse of the rating portal by posting ratings without a real background of experience, the defendant is basically moving within its area of preserving the functionality of the portal, which is protected by Article 5 (1) sentence 1 of the Basic Law. In view of the overriding interest in protecting consumers from non-authentic reviews (cf. the Bundeskartellamt's sector enquiry on user reviews quoted by the defendant in its response to the complaint), it can even be assumed that the defendant is obliged to inform consumers of suspected cases of manipulation. It is true that the rules on reporting suspicions have been developed by case law primarily with the aim of preventing the press from being unduly restricted in its social mandate to report on contemporary events, including in particular criminal offences (Rodenbeck, NJW 2018, 1227). Like the press, however, the defendant is also dependent on providing information about facts that have not yet been fully clarified or proven, especially since the defendant is not involved in the evaluations and any treatment contacts that may underlie them. The principles of suspicion reporting provide concrete and reliable criteria for dealing with such situations. It is therefore logical to measure the plaintiff's warning against the principles of suspicion reporting.

bb) The prerequisites of the principles of suspicion reporting are as follows:

In any event, a minimum amount of evidence is required which speaks for the truthfulness of the information and thus gives it "public value". The report must not contain any prejudgement of the person concerned; it must not, therefore, give the incorrect impression by means of a prejudicial presentation that the person concerned has already been convicted of the act for which he is accused. The person concerned must also be asked to give his or her opinion on a regular basis before publication. Finally, it must be a matter of serious importance, the communication of which is justified by a general need for information (cf. only BGH, ruling of 12 April 2016, VI ZR 505/14).

At the same time, the Senate - in contrast to the argumentation of the plaintiff's representative in the oral proceedings - also sees no reason to modify or tighten the requirements for reporting suspicions, e.g. through the requirement of proportionality or necessity. In the opinion of the Senate, the conditions set out above are suitable to take sufficient account of the conflict and the clash of conflicting interests in the case of suspicion.

(1) There is a minimum amount of evidence to justify the warning notice.

(a) There is a minimum stock of evidence for the existence of purchased / manipulated evaluations in the plaintiff's profile.

The Regional Court considered it plausible that the defendant had found out from emails and IP addresses that evaluators were working for evaluation providers and that these evaluators were also supposed to have evaluated the plaintiff's medical profile. The fact that those users submitted purchased reviews was admitted by other doctors who had been rated by those users.

In the first place, the applicant rightly challenges those findings by stating, first, that the defendant merely stated that doctors admitted that they had purchased reviews, but not that the user specifically listed had published a false review.

Notwithstanding that, the defendant has provided a minimum amount of credible evidence for the suspicion of manipulated / purchased reviews.

First of all, the defendant has shown that individual users who (also) rated the plaintiff have rated the same combinations of doctors, which indicates that they have processed the rating orders made available to them. For example, four evaluators who evaluated the plaintiff also evaluated a doctor E from Stadt3 (cf. 2a b, c, d; note: figures refer to the presentation in the statement of defence); four evaluators also evaluated a doctor E from Stadt1 (cf. 2d, e, f and g). Two of these evaluators also evaluated doctors F from town4 and B from town1 (cf. 2c and g). Two further evaluators also evaluated a doctor G from town5 in addition to the plaintiff (cf. 2j and 2k). These combinations of doctors are already striking, especially as the doctors also practice at different locations, some of which are far apart.

In addition, some of the doctors who were rated by users who also rated the plaintiff admitted to having bought ratings, such as doctor E from town3 , doctor L from town6 (2h), doctor I from town7 (2i), doctor G from town5 and - according to the defendant's new submission in the statement of defence - also doctor B from town1. Some of the doctors denied that they had been involved in the purchase themselves, but did not take action against the warning. Furthermore, the defendant submitted that the profile of doctor E from Stadt2, who was evaluated by a user who also evaluated the plaintiff, could be linked to the testerjob.de portal. In the statement of defence, the defendant also submitted that doctor B from Stadt1 had provided it with a screenshot from his customer account with GoldStar Marketing, in which users appeared whose data corresponded to ratings in 2c and g.

These statements, which were supported by screenshots from the doctors' ratings, were also made credible by the - somewhat more general - affidavit of employee J (pg. 115 of the German version).

In the overall view, these circumstances represent a minimum amount of evidence for the assumption that purchased or manipulated evaluations are or were found on the plaintiff's profile.

In the view of the Landgericht, it is irrelevant in this connection whether the plaintiff was able to assign one of the assessments classified by the defendant as manipulated to a ... patient or whether the doubts which the defendant asserts also predominate in this respect. The fact remains that there are still a large number of assessments in respect of which there is sufficient evidence of manipulation.

(b) The suspicion of possible manipulation falls first on the plaintiff. As the profile holder, he is the one who benefits directly from possible manipulated positive evaluations. Therefore, it is basically up to the plaintiff to clear up the accusations or to cooperate in clarifying how the alleged manipulations occurred. The plaintiff has not sufficiently complied with this.

(aa) Without success, the plaintiff puts the assessments in the context of alleged attempts at blackmail.

In that regard, the plaintiff complains that the Regional Court did not take sufficient account of the fact that he had provided prima facie evidence of having repeatedly received letters from blackmailers threatening to pass on to the defendant the data of the positive evaluations already published if he did not pay € 500.

The applicant's arguments concerning alleged blackmail attempts are already partly contradictory. In a lawyer's letter of 18 February 2020 (Annex MK3, p. 32 of the annex), in which he rejected the accusation of buying or manipulating ratings, the plaintiff first stated that he was aware that criminals had recently been writing to doctors and threatening to publish falsified ratings in order to blackmail doctors into deleting the ratings only after a certain amount had been paid to the fraudsters. It is not clear from this that he himself had received such a letter, which, moreover, would only make sense in the case of negative evaluations. In the application (p. 7, sheet 10 of the application), the plaintiff first repeated this statement and then stated that there was a concrete threat to publish positive evaluations (note: italics by signatories) and to forward the relevant data, in particular the IP addresses, to the defendant. At the same time, the plaintiff claimed for the first time to have received such threats on several occasions, in which he was informed that, in addition, already published (italicised by signatories) ratings would be reported to Bewertungsportal1 if he did not pay. In his affidavit of 22 April 2020 (p. 39 of the German version), the plaintiff affirmed that he had repeatedly received letters from blackmailers threatening to forward the data of already published positive manipulation ratings to Bewertungsportal1 if he did not pay €500. However, the plaintiff was not able to submit such letters. Finally, in a letter dated 28 May 2020, the plaintiff added that it concerned approximately 10 emails sent to his old email address a few months ago, which he rarely uses. They had ended up in his spam folder. He had partially deleted the emails after reading them. Moreover, the spam folder regularly empties itself.

Moreover, this submission - even if it is partly under oath - is not plausible. It is already not clear why the blackmailers should take the "diversion" via the threat of good ratings, which then indirectly lead to the incriminating warning, instead of - which is more obvious - threatening with negative ratings. The blackmailers would then have had to rely on the reaction of the defendant, which cannot be predicted with certainty, as the Regional Court correctly points out. The consideration of the Regional Court that it is implausible that the blackmailers did not contact the plaintiff again after the positive evaluations had been stopped in order to try to obtain money again is also correct. Moreover, the defendant has provided a sworn affidavit (BI. 115 of the Statute of limitations) to show that no doctor has yet approached her for such blackmailing. This also speaks against the plaintiff's submission, especially since it is very unlikely that he alone is supposed to have received such letters. Finally, it is also unlikely that, assuming the existence of the threatening letters to the plaintiff, the manipulated evaluations are based on them. According to the defendant, she did not - as threatened in the alleged blackmail letter - become aware of the plaintiff by means of an external reference, but because of conspicuous features in other doctor profiles which showed the above-mentioned parallels to the plaintiff's profile.

Without success, the plaintiff also refers to an e-mail submitted by and addressed to the defendant (p. 92 of the annex) by an anonymous member of a product and evaluation portal, in which the member complains that he was offered money for the incorrect evaluation of doctor N from Stadt8 . Contrary to the plaintiff's view, this cannot be regarded as a denunciation of a doctor by a blackmailer. Firstly, doctor N has in the meantime conceded the purchase of evaluations. On the other hand, this e-mail - unlike the alleged blackmail attempts - does not tie in with an evaluation that has already taken place, but in it the defendant is just informed that such a desired evaluation will not take place.

Finally, the alleged attempts at blackmail cannot be supported by the argument that it is sufficiently known that "positive" backlink structures have been built up for competitors in the competition for a long time, knowing full well that algorithms of search engines such as google see backlinks bought in them and punish the websites (see page 186 of the German version). For it is not obvious that these backlink structures are associated with an attempt at blackmail.

(bb) Without success, the plaintiff asserts that, due to missing information in the ratings, he has no further indications as to which patients the ratings came from; this is also no longer comprehensible due to the passage of time.

After all, the plaintiff was able to assign an objectionable assessment to a patient (the "..."), even if the defendant points out some peculiarities in this respect. This shows that an assignment may well be possible if treatment has actually taken place. Moreover, the evaluations are linked to a date and are sometimes very concrete (e.g. evaluations 2e and g).

(cc) Also without success, the plaintiff refers to the fact that he did not need to make a purchase because of the "top placement" option already taken up. This option has no influence on the grading.

(dd) Finally, a possible liability of the plaintiff has not been eliminated on the basis of the affidavits submitted by the plaintiff on 22 April 2020 and 26 June 2020, in which the plaintiff fully denies any liability.

The Senate is aware that the plaintiff made the statements in full knowledge of the criminal liability of false affidavits. Nevertheless, the Senate is of the opinion that they do not remove the plaintiff's possible responsibility, as doubts remain as to how the profile of the plaintiff was manipulated. In this respect, it must also be taken into account that the plaintiff attempted to exculpate himself with the - contradictory - reference to blackmail attempts and limited himself to denying his own responsibility as well as the responsibility of any person from his environment (cf. e.g. the lawyer's letter of 27 February 2020, Annex MK4, p. 34 of the annex), the latter without making any concrete efforts to clarify the matter. Contrary to the opinion of the plaintiff, further research on his part is conceivable. For example, he could have asked the patients he had treated in the run-up to the evaluations whether and who had given the positive evaluations in question on Bewertungsportal1. For the first time, the plaintiff made a statement on 13 July 2020 regarding enquiries in his environment, whereby it remains open, for example, whether he questioned the formerly employed ...physicians or his wife about the events at all. For this reason, the Senate has doubts about the plaintiff's possible responsibility, which could not be invalidated by a personal hearing, as the plaintiff had asked for a release from the initial order to appear in person at the specially scheduled hearing.

(2) Contrary to the plaintiff's view, the warning notice is not prejudicial.

(a) The design in itself is not objectionable. The warning notice is not to be seen in the list of results as such. Only when you click on the profile in the results list and it opens does a small red exclamation mark appear in the upper left corner of the overall grade. The textual warning notice, in turn, only opens when the cursor is moved over it - according to the plaintiff's presentation - or when it is clicked on - according to the Senate's self-experiment.

(b) Even the text as such does not fall under the verdict of prejudgement.

Admittedly, the Senate does not consider it a successful formulation when it states that the profile holder has been "confronted" with the facts of the case and that it cannot be assessed "definitively at present" whether the manipulation was initiated by the profile holder. The text also alternates between "suspected cases of manipulation" and "manipulations". Nevertheless, the Senate still considers the text to be admissible.

An informed reader can conclude from the wording that it is only a matter of suspicion of manipulation initiated by the plaintiff. There is talk of "suspected cases of manipulation" and of the fact that it is not possible to conclusively assess whether the profile owner is responsible for the manipulations. At another point, the defendant points out that the matter has not yet been clarified. This makes it clear in several places that the defendant has not yet been able to locate and eliminate the origin of the manipulations. It must also be taken into account that this is a standardised text. For example, the passage "In order to enable users to get an idea of the credibility of the ratings of a profile, we mark profiles in which suspected cases of manipulation in the form of purchased or unfairly influenced ratings have occurred" makes it clear that all profiles in which abnormalities have occurred are marked. In combination with the reference to the previous inexplicability of the origin, an informed reader will thus be able to conclude that only anomalies have been established to date, but not the responsibility of the plaintiff. The meaning of the warning notice is also apparent to the reader from this passage. As the warning states that the labelling is intended to ensure an undistorted opinion ("So that users can form an impression of the credibility of the ratings of a profile"), it is clear to the understanding recipient that the warning is not primarily about apportioning blame, but rather about warning and protecting the user of the platform against a possibly non-authentic opinion. However, this purpose is also fulfilled if the profile owner is not responsible for the manipulations. This speaks against a prejudicial character of the reference.

Contrary to the plaintiff's opinion, the third paragraph does not imply anything else either. The reference to the fact that it cannot be ruled out that manipulated ratings are still present on the site is, against the background that the inexplicability has been pointed out several times before, only to be understood as meaning that the opinion resulting from the ratings may still be distorted. No statement about responsibility can be inferred from this statement.

In addition, the Senate pointed out in the appointment that a search had shown that the warning notice currently reads differently. The defendant has - probably in view of the ruling of the Regional Court Stadt2 - added the sentence "In comparable cases, it has occurred that the manipulations were initiated without the knowledge of the profile owner by colleagues in the practice, employees, family members or agencies involved. Without this being a decisive factor in the present case, the Senate considers this version to be more balanced and suggests in principle that it should be retained.

(3) Without success, the plaintiff complains that he was not sufficiently heard before the warning notice was published. The Regional Court explained in a comprehensible manner why it was sufficient for the defendant to refer for the purpose of the hearing first to two evaluations considered to be exemplary. The plaintiff did not raise any specific objection to this.

Nor is the Senate able to follow the plaintiff's argument that the defendant should have included his opinion in the warning notice.

Although it is generally recognised that the essential content of a statement by the person concerned must be reproduced to an extent appropriate to the scope of the report (MüKo BGB/Rixecker Annex to § 12 marginal no. 206), the Senate does not agree with the plaintiff's argumentation that the defendant should have included his statement in the warning notice. Irrespective of the fact that it is obviously a pre-formulated standard warning notice, the defendant, however, has not omitted any facts which could concretely exonerate the plaintiff. In particular, in view of the inconsistencies described above, there was also no reason to include a reference to possible attempts at blackmail.

(4) Finally, there is also a public interest in the warning label.

(a) A public interest in information results from the fact that the defendant, as an assessment platform, assumes a socially desirable function as an information mediator and thus makes the exchange of opinions and reports on treatment experiences possible in the first place (cf. BGH, judgement of 20 February 2018, ibid.; judgement of 14 January 2020, ibid.) For the exchange of opinions made possible in this way, it is of great importance that the subjective experience reports expressed there are based on actual experience. Only then can an evaluation platform fulfil its purpose of supra-regional exchange of experience reports and opinions. If the image of opinion is distorted in favour of a doctor by purchased evaluations, the credibility of the platform and with it the possibility of exchanging opinions suffers. Users of the platform - 6 million per month according to undisputed information from the defendant - therefore have a considerable interest in knowing whether the experience reports are those written by real patients or whether they were possibly bought in addition. A doctor's profile on the defendant's platform can only become a suitable basis for the choice of a doctor if the reports are genuine experience reports. If there is any doubt about the authenticity of the testimonials, every potential user of the platform - this is anyone without prior registration - has an interest in being informed about them. This applies regardless of the fact that the ratings given by (predominantly) medical laypersons may only allow a few shots at the doctor's professional qualification. This is because the decision on which doctor to consult also takes interpersonal factors into account.

Another argument in favour of a public interest in information is that, in the opinion of the BGH, the above-mentioned socially desirable function of operating a rating portal also includes the control of user contributions, for example with regard to the misuse of the portal through multiple ratings or ratings without an actual background of experience (BGH, ruling of 14 January 2020, ibid.). However, if the control of user contributions by the portal operator is already in the public interest, this must also apply to the communication of the result of this control.

The public interest in information already exists when manipulation is suspected. For as long as it is not clarified who is responsible for the manipulation, the danger of a distorted opinion continues to exist even if conspicuous contributions are deleted. Against this background, it would not be sufficient - but this was the opinion of the plaintiff's representative in the oral hearing - to merely delete possibly manipulated ratings, but to dispense with a warning notice.

(b) Contrary to the plaintiff's opinion, a public interest in the warning notice cannot be denied either because - according to the plaintiff - the disputed evaluations originated in 2019, but the notice has been published (only) since April 2020 and is not limited in time.

On the one hand, the last evaluation listed and objected to by the defendant in its statement of defence dates from 5 February 2020 (cf. no. 2h), so that the warning notice published from the beginning of April 2020 is in a temporal context to the suspected cases of manipulation. On the other hand, the very fact that the objected assessments originate from a period from mid-May 2019 to the beginning of February 2020 makes it clear that the time component is important in several respects. For example, the problem with possibly manipulated evaluations is that they are not easily discovered, but rather it can take a long time for suspected manipulations to become apparent in combination with evaluations on other profiles and investigations of the defendants. It is also apparent that, as the defendant has also pointed out, suspected manipulations are not usually introduced into a profile all at once, but gradually. This also justifies keeping the warning for a certain period of time to monitor and ensure that no further suspicions are raised. In this respect, it is unsatisfactory for the plaintiff that there is no concrete time limit and that the defendant has only stated in general terms when a warning notice is no longer required. However, the Senate considers the previous period of a good six months to be still justifiable in any case. However, the defendant will now be required to review the continued appropriateness of the warning notice and to provide the plaintiff with concrete reasons or suspected cases if it still considers the warning notice necessary.

In the light of the above, the immediate appeal will not be successful.

C.

The decision on costs follows from § 97 (1) ZPO.

The determination of the amount in dispute shall be based on the determination at first instance.