At our recent panel discussion on AI in advertising at the AI Agenda, Lewis Silkin partner Geraint Lloyd-Taylor was joined by Guy Parker, Chief Executive of the Advertising Standards Authority (ASA), and Hayley Fletcher, a Senior Director for Consumer Protection at the Competition and Markets Authority (CMA), to discuss how regulators are responding to the rapid adoption of AI across advertising and marketing.
Key takeaway
The clearest message from the discussion was that UK regulators do not see AI as creating an entirely new category of legal risk. Instead, existing advertising and consumer protection principles continue to apply. If you adopt AI in advertising, you need to focus less on the technology itself and more on the outcomes it produces, making sure that claims are accurate, consumers are treated fairly, and appropriate human oversight remains in place. In an era of expanded CMA powers and increasing regulatory scrutiny, accountability remains firmly with the advertiser, regardless of how the content was created.
Let’s take a closer look at what the panel discussed
Are regulators keeping pace with AI?
Asked about the key challenges of the past 12 months and whether regulation is keeping pace with technological change, Guy said that the existing UK advertising framework is well equipped to address many AI-related issues.
He noted that the CAP and BCAP Codes are principles-based and so do not require AI-specific provisions to regulate AI-generated content. He said that a key issue is transparency and disclosure. While there is growing interest in whether businesses should disclose the use of AI in advertising, targeting and content creation, Guy cautioned against simplistic solutions. Requiring an AI label on every advertisement could have unintended consequences and may not always provide meaningful information to consumers.
Hayley struck a similar note from the CMA's perspective. Consumer protection law focuses on whether claims made to consumers are fair and accurate. Whether those claims are generated by a human or by AI is largely irrelevant. What matters is the impression created and the impact on consumers.
She emphasised that the Digital Markets, Competition and Consumers Act 2024 (DMCC Act) has significantly strengthened the CMA's enforcement powers, making it more important than ever for businesses to be intentional about how they deploy AI. Maintaining consumer trust is critical, particularly given the reputational and financial costs that can arise when it is lost – it is not just about the fines.
Regulators are using AI too
The panellists noted that regulators themselves are increasingly making use of AI technology. Guy highlighted that the ASA uses AI tools to help monitor advertising and identify potential compliance issues.
Hayley stressed that the CMA's approach is not to hinder innovation. It has focused on providing businesses with guidance, including its consumer protection and AI guidance issued earlier this year. She noted that helping businesses comply at an early stage is preferable to addressing consumer harm after it has occurred.
When should businesses disclose the use of AI?
The discussion turned to several recent ASA decisions, including the Woofy and ShroomIQ rulings. Guy suggested that businesses should think carefully about the purpose of any AI disclosure. The key question is what risk or potential consumer misunderstanding the disclosure is intended to address. For example, if an advertisement includes obviously fictional elements such as dragons, consumers are unlikely to need an AI label to understand that the imagery is not real.
At the same time, he warned that an AI label is not a "get out of jail free card". Simply disclosing the use of AI will not cure a misleading claim. Transparency measures must be accompanied by compliance with the substantive advertising rules.
One area where disclosure will be particularly important is the use of AI-generated influencers. Consumers should be informed when an influencer is not a real person, and the CMA is also interested in issues relating to fake influencers and the impressions they create.
Human accountability remains essential
A recurring theme throughout the discussion was that businesses cannot delegate compliance responsibilities to AI systems. Hayley emphasised that companies should focus on what they are communicating to consumers and whether the communication is likely to influence consumers' transactional decisions. If AI technology is used to make products, services or results appear better than they really are, regulatory risks are likely to arise. Many issues emerge not because AI is used, but because AI-generated content is not properly reviewed before publication. Brand owners remain responsible for the claims made in their advertising.
The panel highlighted the ASA's ruling against BCCR BeliefCoding as an example. Among other issues, the advertisement included a claim that 83% of users had seen improvement. AI had apparently been used to summarise an underlying paper and generated an inaccurate summary, contributing to the ASA's concerns about the evidence supporting the claim.
Both regulators were clear that businesses cannot easily contract out of their consumer protection obligations simply because AI played a role in producing the content.
Are there regulatory gaps?
Asked whether there are gaps in the current regulatory framework, Guy suggested that the UK already has a substantial body of regulation, although some elements of the wider regulatory landscape are still bedding in. He observed that regulators increasingly share common priorities and work closely together.
Guy also referred to a recent ASA ruling concerning AI-generated video content and synthetic media, including cases where static images are converted into videos which objectify women. A concern is that many problematic advertisements originate from unidentified or difficult-to-trace actors, making enforcement more challenging.
Hyper-personalisation and targeted pricing
The panel also discussed AI-driven targeting and personalised pricing. Hayley acknowledged that the CMA seeks to anticipate emerging technological developments but cannot always predict exactly how new technologies will be deployed. Nevertheless, she reiterated that the same underlying legal principles apply regardless of whether AI is involved.
When assessing practices such as personalised pricing, drip pricing, refund processes or AI-enabled customer service, the key issue remains whether consumers are being treated fairly.
She noted that differentiated pricing can be lawful. While regulators scrutinise such practices closely, they may be permissible where they are relevant to consumers and used responsibly.
The discussion also touched on regulatory cooperation. Guy explained that the CMA is not a formal "backstop" for the ASA, but the organisations work closely together alongside Trading Standards and sector-specific regulators. Decisions about enforcement action are often coordinated, with regulators deciding who is best placed to act in a particular case. Secondary ticketing was cited as an example of this collaborative approach.
Looking ahead
In response to a question about whether the EU AI Act has influenced the ASA's approach, Guy noted that the Government is currently considering issues around AI labelling, an area that is likely to remain under close scrutiny.
The session concluded with a discussion of complaints trends. Guy explained that the ASA receives around 40,000 complaints each year and is seeing increasing evidence that consumers are using AI tools to help draft complaints. Interestingly, he suggested that this may be improving complaint quality. While AI-assisted complaints are often longer and sometimes adopt US-style language, they can be more focused and accurate. In some instances, consumers reportedly use AI to assess a concern and conclude that the issue does not in fact warrant a complaint to the ASA.
Explore more from AI Agenda 2026
On 16 September 2026, we hosted AI Agenda 2026 for senior leaders, legal experts and innovators exploring what it takes to deploy AI safely, strategically and at scale. Explore all the insights from other sessions - covering topics from agentic AI and governance to IP, procurement, litigation and leadership.
