The Advertising Standards Authority (ASA) received 37,284 complaints last year, involving 24,015 ads.
From those, the watchdog secured the amendment or withdrawal of 33,903 adverts, up from 27,378 in 2023.
Online was the most complained about media, with 20,423 complaints, relating to 16,530 ads. While TV was in second place with 12,383 complaints, relating to 4,008 ads.
Although the ASA is best known for investigating complaints, it is now able to review millions of online ads every month, thanks to its AI-powered Active Ad Monitoring system.
Subscribe to Marketing Beat for free
Sign up here to get the latest agency-related news sent straight to your inbox each morning
Using AI
According to new figures published in the ASA and Committees of Advertising Practice (CAP) 2024 Annual Report, the watchdog assessed the suitability of 28 million online adverts last year.
To put that figure into context, by the end of the second half of 2023 around half a million ads were processed by the automated system, but by the end of the same period in 2024, it processed more than 3 million.
The outcomes are equally significant. Of the 33,903 ads which were amended or withdrawn, 94% were identified using the Active Ad Monitoring system.
The vast majority (84%) were non-paid ads online, and most related to claims on companies’ own websites or social media accounts.
The ASA’s six-strong data science team is headed by Adam Davison, and writing in the annual report, he said they improved their “AI-based toolkit for monitoring online advertising”, during 2024.
“As a result, it processed 28 million ads in the last year, giving us much greater visibility of ads appearing online,” he said.
The Active Ad Monitoring system was also the source of 41 published formal rulings last year, almost 15% of the ASA’s total output.
The technology uses Large Language Models (LLMs) to speed up the review of content, and the system will be further developed to increase the number of topics it can monitor.
“We are also actively experimenting with how these tools may make our internal processes more efficient in the future,” Davison said.
“We will continue to integrate it more deeply into our work, ensuring it is not just innovative but also leads to us delivering visibly better regulation.”
High-priority projects
As well as being able to review huge numbers of online ads, the ASA also uses AI-based tools in most of its “high-priority projects”. For example, around climate change and the environment, influencer marketing, financial advertising, prescription-only medicines, gambling, e-cigarettes and cosmetic surgery clinics based outside the UK.
Miles Lockwood, the ASA’s director of complaints and investigations, wrote in the report that the Active Ad Monitoring system had identified that many airlines had made misleading online claims that suggested flying was ‘sustainable’ and ‘eco-friendly’, “where the evidence was lacking”.
One ad the ASA ruled was misleading, was by Air France (left) as the airline did not provide any evidence about how they were protecting the environment or making aviation sustainable.
“Following a series of rulings and engagement with the sector, we’ve been pleased to see improvement and our routine monitoring now shows high levels of compliance,” Lockwood said.
“Of the circa 140,000 ads monitored in April 2024, five were found to be non-compliant.”
Protecting children
Another area of ongoing investigation for the ASA is around age-restricted ads and protecting children from online harm.
In March 2024 the watchdog launched two “groundbreaking projects”, aimed at better understanding where responsibility lies for inappropriately targeted or irresponsible ads that appear online.
One project employed innovative technology to monitor for ads for age-restricted products, including alcohol, gambling and foods high in fat, salt or sugar (HFSS), on sites of particular interest to under-18s.
The other used the same approach to monitor for seriously offensive and potentially harmful ads, which appeared in mobile quiz and game apps.
Writing in the annual report, Emily Henwood, ASA operations manager for complaints, said the watchdog has also continued its work on vaping ads that appear on social media “where, “under the law and our rules, they shouldn’t”.
“Our monitoring revealed problematic content from both influencers and businesses on all social media platforms,” she said.
The ASA also issued its second enforcement notice, relating to E-cigarettes, to “hundreds of key players” said Henwood, including brands and retailers, “directing them to stop any advertising on social media”, including retailers’ own organic ads, paid ads, and influencer posts.
“Compliance action is ongoing and involves having social media accounts taken down where an advertiser appears unwilling or unable to stick to the rules,” she said.
Transparency and clarity
The ASA also took action last year to protect vulnerable people from misleading, harmful or offensive ads.
They included adverts with unsubstantiated health claims, such as one for Feminapause (left).
The complaint was upheld because food supplements cannot imply that they treat or cure the symptoms of menopause. The ad also “falsely implied Feminapause had been approved by the Food Standards Agency (FSA).
The ASA also investigated complaints about eight different drug and alcohol rehabilitation clinic referral services. Again, AI was used to identify many of the adverts.
“The issues were transparency and clarity, as the ads were likely to mislead and could affect particularly vulnerable people and their families,” Jo Poots, the ASA’s head of complaints and investigations, wrote in the annual report.
“As well as acting on complaints, we used our Active Ad Monitoring system to proactively search for online ads that might break the rules.”
Poots said the ASA also “shone a light on problem health claims” by ruling against a range of supplement ads that claimed to treat anxiety, ADHD and autism, as well as ads that claimed CBD was suitable for consumption by children.
