Clifford Chance

In today's fast growing world of advertising, staying ahead of the curve is essential. There are many way businesses can look to do that, but one that is growing in popularity is using generative AI. How this will work in practice is still up for debate – and will be an interesting area to watch over the next few years, particularly as the global debate on whether there's copyright in the AI-generated works (and if so, who owns it) rumbles on. However, CAP and the ASA have recently released guidance on the use of generative AI in advertising campaigns, to make sure that advertisers stay on the right side of the law.

What is Generative AI?

In a nutshell, generative AI is a type of artificial intelligence that can create original content for you, based on the user's request. From images to text to music, it draws from a vast pool of existing data to produce new works for use (see our articles:  AI-Generated Music and Copyright  and What's wrong with this picture? AI-created images and content: copyright ownership and infringement ). As the ASA notes, AI has the potential to revolutionise ad campaigns, creating tailored ads that are personalised to each consumer, generating a year's worth of branding content with just a few clicks.

Will the ASA treat AI-generated content differently from human-generated content?

In short- no. The ASA has advised that if the ad falls into its scope, regardless of how the ad was created, the CAP Code rules will apply (or BCAP Code for TV or radio ads). This means that the ad still has to comply with all the regular requirements, including that the claims are not misleading, and they have to be true and substantiated. This is because the consumers' interpretation is key, and that doesn't change based on how the content itself was generated.

While the ASA are not aware of any ruling they have made that involved AI generated works, it is certainly something they will face in the near future. Their recent advice on using AI generated content highlighted the following points:

AI Generated Images still need to be accurate

While the method of content creation may not change the ASA's guidance, certain scenarios warrant attention. For ads using AI-generated images or videos to portray a product, accuracy in reflecting the product, its use or its efficacy is paramount. Remember, these still need to be truthful, and can't create a misleading impression.

Guard against biases amplified by AI models

Another area of concern is generative AI created content that exacerbates harmful biases. This could lead to ads that are socially irresponsible, breaching rules on harmful gender or other harmful stereotypes.

Recent examples have shown that AI generated content may tend to amplify biases present in the data they've been trained with – the examples cited by CAP include specific generative AI tools that tended to portray men or individuals with lighter skin tones in higher-paying occupations, or portraying idealised body standards.

This means that it is important to ensure your content is still socially responsible, and steers clear of harmful stereotypes. If the content isn't sense-checked for these biases, it could lead to ads that are ultimately socially irresponsible. Advertisers venturing into the realm of AI-generated images should be cognisant of this risk and take remedial measures to avoid inadvertently perpetuating harmful stereotypes.

Advertisers are still responsible

The ASA recent Stripe & Stare Ltd ruling confirmed the position that even if the marketing campaign is entirely automated, advertisers bear the primary responsibility for ensuring compliance, at all stages of the ad creation and distribution process. There still is no substitute for a pair of human eyes.

Key takeaways

• Seize the potential: Generative AI offers unparalleled opportunities for dynamic and personalised content creation.
• Compliance is key: The CAP Code or BCAP Code rules still apply, and even AI-generated content needs to be accurate and transparent.
• Truth in Imagery: AI-generated images must accurately represent product efficacy, just like any other visual content.
• Champion responsibility: Make sure your advertising is socially responsible, and be aware of biases and stereotypes.
• Your role: Advertisers are still responsible for compliance, even in automated campaigns.

The integration of generative AI in advertising holds major potential, but it comes with a set of responsibilities and potential risk. Advertisers must approach this technology with vigilance, maintaining transparency, accuracy and ethical considerations.

For personalised guidance on integrating AI-technology into your next marketing campaign, please feel free to contact us.

To keep abreast of key ASA rulings subscribe to The Advertising Brief our quarterly newsletter delivering concise summaries of the most interesting cases from the ASA with our key takeaways for all brand owners to be aware of.