The Advertising Standards Authority (ASA) has recently published an update on its approach to the use of artificial intelligence in advertising. This explains how the ASA is harnessing AI technology through its Active Ad Monitoring system to enhance the efficiency of its own surveillance of ad compliance.
Alongside this, the ASA has issued a report on the use by brands of "AI" as a marketing term, providing guidance for businesses on responsible advertising practices.
On the same day that the update was released, the ASA also hosted a webinar titled "Future-proof Advertising: How will AI change advertising and regulation?". The event brought together a panel of industry experts to discuss the evolving landscape.
Below, we distil the essential insights from the ASA:
Active Ad Monitoring system
The ASA's Active Ad Monitoring system has three capabilities: (i) capturing ads from social media; (ii) identifying ads that may be non-compliant with advertising regulation; and (iii) expert review in which experts can browse and search content.
The ASA has been significantly increasing surveillance of ads with AI. It currently scans over 3 million ads a month with AI, compared to the figure noted in its webinar for last year when it was 100,000 ads a month.
In addition, in 2025 the ASA expects 20% of rulings to be sourced from AI based monitoring. The ASA emphasised, however, that none of its final decisions are taken by computers or technology and are ultimately decided with human review.
Report on use by advertisers
The report on use of the term AI by advertisers (which was also discussed in the webinar) details how the ASA identified 16,000 ads that used the term AI (between July and September 2024), describing use of this term as now "widespread". The breakdown for such ads were: (i) 68% related to business to business (B2B) products and services with AI features (such as accountancy software and cybersecurity vendors); (ii) 9% related to consumer products (for example personal tech, cameras, and home appliances); (iii) 13% concerned new “AI-native” products (such as chatbots, photo editing and tutor tools). The remaining 10% were grouped as miscellaneous.
In regard to ads for consumer products, claims fell into two broad categories: 70% were general claims (i.e. AI being present in the product, such as "AI Laptop" or "AI Powered") and the other 30% were specific claims (i.e. claims as to why a product delivers better performance through the use of AI, such as "AI dynamic tone mapping" and "get powerful AI performance"). The ASA has stressed that businesses need to be especially cautious with specific claims as to how they present AI. For example, advertisers must ensure they do not exaggerate, falsely claim or mislead as to what AI can do. In addition, advertisers must be able to substantiate claims with evidence. The ASA has, of course, already provided similar guidance in the context of green claims, and there is therefore an emerging space for the ASA to focus on regulating ads which are "AI washing".
The ASA will continue to monitor the use of "AI" in advertising over the next 6-12 months and may issue further guidance or take additional actions if deemed necessary.