Garrigues Digital_

The emergence of synthetic content has reignited the debate over whether traditional legal mechanisms are sufficient to protect the right to honor. This right is now facing unprecedented challenges that strain the established standards by which it has traditionally been defined.

Organic Law 1/1982, of May 5, 1982, on the civil protection of the right to honor, personal and family privacy and personal portrayal (LOPCDH) refers to social customs to define the scope of protection of honor. Few, if any, legal concepts have seen their meaning change so persistently throughout our history. We have gone from being a society in which the loss of honor was tantamount to civil death, to the digital era in which insults and lies are the norm. In the process, we have stopped effectively protecting a right that lies at the core of human dignity.

Protection of reputation is not a trivial matter, since honor is ultimately linked to truth and truth, in turn, constitutes an essential part of ethics. Kant had already warned that lying is not a minor wrongdoing; it is an ethical breach that erodes human dignity and undermines the trust upon which social life depends. If this statement was undeniable when lies were spread analogically, it is all the more true in an age in which technology enables lies to multiply, amplifies them and makes them viral in a matter of seconds. This is why indolence in the defense of honor degrades society as a whole, since violating honor is ultimately equivalent to consenting to the rule of lies.

Organic Law 1/1982 in the case of synthetic content

In Spain, the civil protection of these rights lies mainly in Organic Law 1/1982, which provides protection from the unlawful infringement of honor, personal and family privacy and personal portrayal. The problem is that this law was designed for a world of photographs, printed publications and traditional media outlets and not for a world in which a person’s appearance can be artificially generated and mass-disseminated in seconds.

Article 7.7 of the LOPCDH considers the attribution of facts or the expression of value judgments through actions or expressions that may in some way harm the dignity of another person, damaging their reputation or eroding their self-esteem to constitute an unlawful encroachment on the right to honor. In the absence of a statutory definition, case law has understood honor as a person’s good reputation which, like fame and esteem, consists in the opinion that society has of an individual. The common denominator of all attacks on or unlawful infringement of the protection of this right is a diminishment of the esteem in which a person is held.

When assessing potential infringements of honor, it should be borne in mind that honor is not an absolute right. It may, at times, have to give way to other fundamental rights such as freedom of expression which protects the free expression of thoughts, ideas, and opinions, encompassing both subjective assessments and value judgments, including harsh criticism. If what is being expressed are opinions, the right to honor will only prevail over freedom of expression if the content in question is insulting or degrading.

However, before reaching the conclusion that a particular statement is insulting or degrading, it should be noted that the Spanish courts have a high degree of tolerance when it comes to allowing the use of offensive language. In this regard, Supreme Court judgment number 6556/2013, of December 30, 2013 states as follows: “Case law establishes that freedom of expression prevails where expressions are used that, although offensive in isolation, when considered in connection with the information to be conveyed or with the political or social context in which the criticism takes place, reduce their offensive meaning and suggest that an increase in the degree of tolerance is required, even if they may not be entirely justifiable (Article 2.1 of the LPDH states that social customs mark the boundaries of the civil protection of the right to honor). (…). However, these considerations should not be strictly limited to the political arena, since case law applies the exact same principles to situations of tension or conflict in the workplace, labor unions, sports, and procedural contexts, among others.”

Several months ago, initially with the intention of addressing this situation, the Council of Ministers approved the Preliminary draft organic law on civil protection of the right to honor, personal and family privacy and personal portrayal. This law expressly provides that deepfakes can unlawfully infringe personality rights and also includes criteria to assess the severity of the non-pecuniary damage caused. However, regulating fundamental rights is extremely complex, as the Constitutional Court itself has recognized, insisting on the need to address each matter individually, case-by-case (Constitutional Court judgment 46/2022, of February 25, 2022; Constitutional Court judgment 52/2002, of February 25, 2002; Constitutional Court judgment 180/1999, of October 11, 1999).

In this area, as in many others, more legislation does not mean better protection. What is needed is effective judicial protection to assess these rights quickly and swiftly, since without the relevant procedural support, legislative reform of this magnitude will not be sufficient.

European regulatory framework: The Artificial Intelligence and Digital Services Acts

The response to synthetic content does not end with Spanish law. The European framework introduces two relevant laws, namely the Artificial Intelligence Act (AI Act) and the Digital Services Act (DSA), which become operative at different stages of the problem.

The AI Act mainly regulates transparency and the detection of synthetic content, imposing certain obligations for artificially generated or manipulated content to be identified as such (article 50). However, the fact that certain artificial content is detectable does not mean that it is not damaging, because although transparency can reduce the risk of deception, it does not provide comprehensive protection for the person affected.

The DSA in turn mainly comes into play in the dissemination of content by online platforms and social media. The legislative policy of the DSA does not impose a general obligation on online platforms and social media to monitor or validate ex ante, the content uploaded by users. This is both because it is highly impractical (think of the number of videos and texts that are uploaded to social media each second) and also in view of the risk of turning them into arbiters of truth.

Social media and platforms that allow third parties to publish content do not hold editorial responsibility for all the content that circulates via their services. The provider of these services benefits from exemption from liability, for example where a user has uploaded defamatory content, provided that it does not have actual knowledge of the illegal content, or, upon obtaining actual knowledge acts promptly to remove or disable access to the information.

Actual knowledge arises where the content is manifestly illegal and its removal is requested, or following an order from a competent authority identifying the content and requiring that action be taken in that respect. From then on the intermediary can no longer rely on mere technical passivity and must act diligently, removing the content or disabling access to it with the appropriate promptness.

This requirement to remove or disable access to content where it is manifestly illegal or following an order from a competent authority, poses particular difficulties in defamation matters. As opposed to what may occur in some cases of infringement of intellectual property rights, where the illegal nature of the content can be more objectifiable based on ownership of the right or on the unauthorized reproduction of a protected work, sign or subject matter, infringement of the right to honor tends to requires a much more complex contextual assessment.

The platform or social media service is unlikely to be in a position to conclude alone, with certainty, whether a statement is manifestly illegal, is covered by freedom of expression or information, is based on truthful facts, or whether it is defamatory. As a result, removing or disabling access to the content in this area without a court order has greater risks and difficulties attached, since it involves the need to assess the conflicting fundamental rights and aspects that often can only be adequately addressed in court.

That is, when it is the users themselves who create and upload the content, the platform operates, in principle, as an infrastructure for publication and dissemination, not as the author, or as an editorially responsible party. The platform’s liability is triggered not by publishing the content, but rather by not acting appropriately once its illegal nature has been made clear in the terms required by the regulation and it fails to remove it.

The liability of providers of AI models or systems is a different matter however, because the content has not been created by users, but generated by the model or system itself. In this scenario it is questionable whether these providers can rely on the exemption from liability regimes that the DSA reserves for providers of intermediary services with respect to content uploaded by users.

In short, although the legal system provides mechanisms to respond to attacks on honor, it is advisable to realistically assume that protection is, at present, particularly challenging. The most effective course of action when the content is manifestly illegal is to request its removal or that access to it be disabled through the notice mechanisms made available by social media and the online platforms themselves. Where the illegal activity is not manifest (as is often the case with defamation, where the assessment requires an examination of context, accuracy, freedom of expression and honor), the only alternative is to resort to court proceedings, with all the problems that this entails: response times that are incompatible with the speed of the digital dissemination, uncertainty as to the outcome and, at times, the additional risk of amplifying, through media exposure, the actual content being challenged.

This is why it does not seem realistic to believe that the protection of honor is going to improve substantially by simply changing the law. Real change will depend both on new laws and also on a more effective judicial response, together with indemnification that provides better compensation for the damage caused.

This is further compounded by an increasingly evident paradox: anonymity in the digital environment and data protection that is, at times, excessively expansive. Under the guise of protecting fundamental rights, excessive protection of personal data can make it difficult to identify the party responsible, hamper effective protection of the person affected and make it practically impossible to punish the damage to honor.

Defamation is definitely much easier when the defamer acts under the cover - and also the cowardice - of anonymity. That shield enables some to cause damage deliberately, others to manipulate public opinion and yet others to turn lies into a profitable product, designed to attracts clicks, fuel virality and generate economic gain at the expense of someone else’s reputation.

As early as the 17th century, Lope de Vega - exiled for his defamatory sonnets against his lover’s family - expressed with great clarity the importance of honor and the difficulty involved in redressing the damage caused to it: “It is folly to think that by killing the offender, the offense to the offended is removed.”