Spotlight on: Generative AI in advertising
5 March 2024
Generative AI is reshaping the way content is generated and consumed, with far-reaching implications for the advertising industry.
What is Generative AI and how might it be used by advertisers?
Generative AI is a type of artificial intelligence that enables computers to generate content, including text, images, videos, and interactive experiences. It does this by learning from data and copying the patterns it identifies.
While the industry may use this technology to streamline and automate ad creation, there are concerns about the impact of widespread use of AI generated content. If you are considering using generative AI in your campaigns, take a moment to read though our guidance on code compliance below.
What are the ASA’s rules on use of generative AI in advertising?
The ASA Codes apply to all advertisements, in all media. The Codes are technology-neutral, meaning that regardless of how an advertisement is made, it must comply with the ASA rules. Ultimately, the Complaints Board considers the consumer take-out of an advertisement, and how the advertisement was created may or may not be material.
There are some instances where the way in which an advertisement is created may impact whether it breaches the Codes, and below are our top tips for ensuring AI-generated advertising is socially responsible.
One rule in the Advertising Standards Code of particular importance for AI-generated ads is Rule 2(b) Truthful Presentation:
“Advertisements must not mislead or be likely to mislead, deceive or confuse consumers, abuse their trust or exploit their lack of knowledge. This includes by implication, inaccuracy, ambiguity, exaggeration, unrealistic claim, omission, false representation or otherwise. Obvious hyperbole identifiable as such is not considered to be misleading.”
Advertisers using generative AI must ensure that their advertisements accurately reflect their product or service, and do not inflate or falsely represent the benefits to the consumer. For advocacy advertising, claims made in an advertisement generated by AI must be able to be substantiated, and the advertiser’s opinion clearly distinguished from fact.
There is also a risk that AI-generated content may reflect biases already present in the data the model was trained on, which may lead to these types of content perpetuating negative stereotypes.
Rule 1(c) Decency & Offensiveness, states that:
Advertisements must not contain anything that is indecent, exploitative, or degrading, or likely to cause harm, or serious or widespread offence, or give rise to hostility, contempt, abuse or ridicule.
Advertisers who choose to use AI-generated images must be mindful of the risk of inherent bias, and ensure their final content is socially responsible and does not portray negative stereotypes which may be harmful if perpetuated.
Advertisers must also be aware of the rules around data collection and privacy when creating personalised campaigns. The Advertising Standards Code also includes rules around data collection and consent to use of data, and the consumer’s right to privacy. Rule 1(a) of the Advertising Standards Code states:
Advertisements may only portray or refer to personal information that is publicly available. Other personal information may only be collected and used with the consent of the individual concerned.
In summary, advertisers should be aware that the ASA Codes apply to advertisements created through generative AI and have checks in place across their content creation and placement to ensure the resulting campaigns are socially responsible and not misleading.
