Menu
© 2026 World Compliance Association

18/01/2026

The Government is seeking to combat ‘deepfakes’ by classifying them as a violation of the right to honour

The Council of Ministers has approved a draft bill aimed at curbing images created by artificial intelligence

The Government is seeking to combat ‘deepfakes’ by classifying them as a violation of the right to honour

The Government wishes to protect the rights to honour, privacy and one’s own image, and to bring legislation into line with the 21st century. The case of the videos of the underage girls from Almendralejo, who were stripped naked using artificial intelligence (AI), highlighted the new dangers posed by these tools. The government therefore considers it necessary to reform the legislation so that citizens are not left defenceless. On Tuesday, the Council of Ministers approved the preliminary draft of the organic law on the civil protection of the right to honour, personal and family privacy, and one’s own image, which aims to curb the manipulation of images and even voices. “It is a more protective piece of legislation, adapted to new technologies,” said the Minister and Government Spokesperson, Elma Saiz. The unauthorised use of a person’s image or voice that has been created, simulated or manipulated using AI or similar technologies will be deemed unlawful. This refers to what are known as ‘deepfakes’, which the text terms ‘ultra-impersonations carried out using artificial intelligence’.

The issue has returned to the spotlight in recent weeks following the circulation on social media of sexualised images of women and minors generated by Grok, the AI tool of the social media platform X. Whilst Grok blocks nudity, it is possible to request images of women or minors in bikinis.

Among the key measures set out in the draft bill, to which this newspaper has had access, is raising the age at which minors may give consent for the use of their own image to 16, in line with the provisions of the law on the protection of minors in digital environments. This latter piece of legislation, which is currently being debated in Parliament, also addresses deepfakes, but from a criminal law perspective: it classifies them as a criminal offence when, without the authorisation of the person concerned and with the intention of undermining their moral integrity, these technologies are used to simulate “situations of a sexual nature or that are seriously humiliating”. Prison sentences of one to two years are therefore envisaged in such cases.

However, the regulation now being put forward by the Government focuses on the civil sphere. To calculate the compensation to which those whose rights have been infringed are entitled, the text sets out the criteria for judges to assess the seriousness of such infringements: the social repercussions, whether there has been a repeat offence, the severity of the language used, etc. The court must determine the compensation for non-pecuniary damage, which under no circumstances may be of a symbolic nature. And final judgements will be published in the Official State Gazette (BOE) if requested by the injured party. “It is by no means clear that this ban will lead to higher compensation, because it does not define what is or is not ‘symbolic’,” says Jorge García Herrero, the data protection officer.

The reform also provides that, in the case of those aged 16 and over, an act shall be considered an unlawful intrusion – even if the person has given their consent – where their dignity or reputation is undermined. “That is tantamount to recognising that consent is null and void for these purposes. Drafting provisions in such black-and-white terms, legally prohibiting something without any possibility of exception, always leads to problems,” adds García Herrero.

It will also be possible to prohibit, by means of a will, the future use of one’s own image or voice for commercial or similar purposes after one’s death. The disclosure of details of a crime by the perpetrator, where such disclosure could cause further harm to the victim, shall be considered an unlawful interference with the victim’s rights. This applies, for example, in ‘true crime’ cases.

The Government’s draft legislation provides for exceptions. In addition to retaining those already provided for in the legislation in force to date (Organic Law 1/1982), such as recordings authorised by the competent authority or the dissemination of private communications containing newsworthy content, provided they relate to matters of general interest, new exceptions are added concerning the use of AI. If the person concerned is a public figure, freedom of expression is protected, provided that these techniques are used in a creative, satirical or fictional context and it is clearly stated that this technology is being used.

The draft bill to be approved this Tuesday by the Council of Ministers does not relate solely to minors, but updates, in the civil sphere (rather than as a criminal offence), the protection of the rights to honour, privacy and one’s own image. It therefore replaces the current law, which dates from 1982. For example, the draft bill stipulates that sharing personal images on a social media platform shall under no circumstances entitle third parties to share them on other platforms or distribution channels. For the legal expert Borja Adsuara, a specialist in digital law, the previous legislation already provided sufficient safeguards. “The protection of the right to honour is not being strengthened because it is already enshrined in law. There is no need to tweak the rules every time a new technology emerges. This is merely a publicity stunt, in the wake of the Grok incident, to make it seem as though the Government is doing something,” he says.

The adoption of this legislation forms part of the measures to promote a higher quality of public debate set out in the Action Plan for Democracy, approved by the Council of Ministers last September. As it is an organic law, its approval in Parliament will require an absolute majority, which seems difficult to achieve.

A problem with multiple legal fronts

The authorities have been considering for some time how to tackle the unwanted dissemination of manipulated images. The European Commission has raised a hue and cry and has asked X to retain all its documentation relating to Grok throughout 2026, as part of the ongoing investigation it is conducting into the platform to analyse its policy on illegal content. Governments in various countries have also reacted, ranging from France – which has taken legal action – to Malaysia and Indonesia, which have directly ordered that access to Grok be blocked.

In Spain, the Ministry of Youth and Children last week sent a letter to the Public Prosecutor’s Office expressing its concern over the proliferation of sexualised images of minors and voicing its fear that these may constitute offences relating to child pornography and offences against moral integrity. The Minors’ Unit has received this letter and will proceed to examine it, as a spokesperson explained to this newspaper.

Aware of the risks posed by new AI applications, the Spanish Data Protection Agency (AEPD) published a circular this Tuesday warning of the risks of manipulating images with these tools and pointing out that doing so may even constitute a criminal offence. The agency points out that images – whether photographs or videos – in which a person is identifiable constitute personal data, and this is the case regardless of whether the image is real or generated by AI systems. Therefore, forwarding such images, uploading them to the internet, sharing them on social media or modifying them using AI tools “constitutes the processing of personal data”. And, if the data subject has not given their consent for this to happen, it constitutes a breach of the General Data Protection Regulation (GDPR).

The visible risks that may arise from the use of image-manipulation systems, according to the AEPD, range from the sexualisation and dissemination of intimate content to the attribution of untrue facts or damage to the victim’s personal relationships. There are also hidden risks, such as the effective loss of control over the image, which passes into the hands of an external provider; the retention of that image and the making of copies; or the risk of persistent identification of the individuals concerned, for example through automated facial recognition systems.

“Uploading images or videos of third parties to artificial intelligence systems is not a neutral act, even when there is no subsequent dissemination or intention to cause harm”, concludes the Agency, which also points out that the risks mentioned often fall outside the scope of data protection legislation, “even though other fundamental rights, such as honour, privacy or one’s own image, may be affected, or other provisions of the legal system, including the Criminal Code, may be applicable”.

Source: https://elpais.com/

NEWS LINK

reCAPTCHA protection