AI adverts banned over sexualised images and explicit content
The UK's advertising regulator has banned adverts for AI applications after finding they objectified women and promoted tools capable of creating sexualised material from people's
By BBC News
A series of adverts promoting artificial intelligence applications has been banned by the UK’s advertising regulator over concerns about the sexualisation and objectification of women.
The Advertising Standards Authority ruled against five social media adverts after finding that the material breached advertising rules and was likely to cause serious offence.
The adverts appeared on Meta’s platforms and promoted AI applications capable of generating or manipulating images and videos.
Some encouraged users to upload photographs of people they knew, including their “crush”, before using artificial intelligence to transform the images into sexualised videos.
The regulator was particularly concerned about the potential for this type of technology to be used on images of real people without their consent.
In one case, the ASA found that advertising for an AI companion application created by Animcha Ltd included sexualised imagery involving a girl who was under 18.
The regulator said the presentation was unacceptable and breached advertising standards.
Separate adverts promoting Nexaipic, an AI portrait generation tool, were also found to have broken the rules.
The cases form part of a wider challenge facing regulators as increasingly sophisticated generative AI tools make it easier to create realistic images and videos from existing photographs.
Technology that can alter clothing, appearance or behaviour in photographs can have legitimate applications, but it can also be misused to create sexualised or explicit material without the knowledge or permission of the person depicted.
The ASA said it takes a “zero-tolerance” approach to advertising that sexualises or objectifies women.
It also warned that advertisements encouraging people to create sexually explicit material from photographs could contribute to real-world harm.
Significantly, the offending adverts were detected using the regulator’s own artificial intelligence technology.
The ASA increasingly uses automated monitoring systems to examine large volumes of online advertising and identify material that may breach its rules.
That gives the case an unusual technological dimension, with AI being used by the regulator to identify potentially harmful advertising for other AI products.
Online advertising presents a particular enforcement challenge because campaigns can be created and distributed quickly across social media platforms and potentially reach large audiences before complaints are made.
Automated monitoring allows regulators to identify some problematic advertisements without having to wait for members of the public to report them.
The latest rulings also highlight growing scrutiny of companies marketing generative AI applications directly to consumers.
As image and video generation becomes more realistic, regulators are increasingly having to consider not only what AI applications can produce but also how companies are allowed to promote those capabilities.
The five adverts covered by the latest rulings cannot appear again in the forms investigated.
The ASA has warned advertisers that using sexualised imagery or encouraging the objectification of women to promote AI products will not be tolerated under existing advertising rules.