We all remember the viral Glasgow Willy Wonka Experience disaster in 2024, where promotional images created with AI couldn’t be further from the reality when customers arrived.
The mismatch was a catalyst for widespread criticism, police involvement and the attracting closing in a matter of hours.
AI also raises concerns beyond misleading visuals. Personalised adverts can reach people who should not be targeted, including children who receive age-restricted promotions or individuals shown sensitive content after opting out.
Issues linked to intellectual property and data protection add further uncertainty for teams using automated tools without clear controls.
Here are five steps to avoid falling foul with AI in advertising.
Responsible use does not mean abandoning AI altogether. With clear processes in place, brands can make use of its capabilities while reducing legal and regulatory risk.
-
1. State when AI is involved where needed
Advertising rules require that consumers are not misled and that material information is provided.
The Competition and Markets Authority (CMA) has indicated that people should know when they are dealing with an automated system unless this is already evident.
This applies to AI-generated images, virtual characters and digital influencers. If these appear human at first glance, viewers may assume they are genuine. A short clarification can help avoid confusion.
-
2. Set expectations in contracts
Agreements with agencies and creators should outline how AI may be used, who checks outputs and who is responsible if errors arise.
Clauses covering approvals, warranties about data sources and audit rights can help manage risk.
-
3. Strengthen content checks
All material should be reviewed to ensure that AI-generated output does not give a misleading impression of how a product looks or works.
In some cases, explaining that AI has contributed to the content may prevent misunderstandings.
-
4. Use AI influencers carefully
Digital influencers should be clearly identified if it is not immediately obvious that they are AI-generated rather than human.
If they appear to ‘test’ a product, you will need to consider whether that is possible for a synthetic character; if not, that format may be unsuitable.
-
5. Take extra care with regulated categories
Campaigns for products with age limits or other restrictions need closer scrutiny.
Personalised targeting should be checked manually to reduce the likelihood of unsuitable placements or outreach to vulnerable groups.
The future of AI in advertising
AI now supports tasks from copywriting to image generation and tailored personalisation.
Used carefully, it can bring genuine efficiencies. Yet missteps can lead to a legal, reputational and regulatory minefield.
By setting clear policies, improving content checks, managing the role of virtual influencers and applying particular care to regulated categories, brands can use harness AI responsibly and safely.