AI in marketing: Balancing innovation, ethics, trust, and regulatory requirements
Have you watched the latest Adidas advertisement, created entirely with artificial intelligence (AI) by filmmaker Blair Vermette? This unofficial ad showcases a vibrant “Floral” collection, featuring AI-generated individuals dancing in floral-printed apparel against Japanese-inspired backdrops.
Vermette utilised AI tools such as RunwayML, Midjourney, Adobe Creative Suite, and Topaz to produce this visually captivating piece.
Despite not being an official Adidas campaign, the ad went viral, blurring the lines between reality and AI-generated content, sparking discussions about AI’s future role in advertising.
The ad has garnered mixed reactions from creative professionals. Some praise its visual appeal and innovative use of AI, while others feel it lacks the emotional depth and authenticity that the human touch brings to storytelling.
Viewers are complaining about the AI-generated advertisement, not because it’s AI-generated, but because the products shown don’t actually exist.
While the advertisement and the products depicted are unreal, this ad provides Adidas with valuable insights into emerging trends. It allows them to understand how to adapt its product designs to better-meet market demands, and align with the general population’s preferences.
Coca-Cola has also explored the use of AI to predict trends and suggest new product ideas based on unmet market needs. Nike, a competitor of Adidas, also uses AI tools to design custom sportswear, understanding its market demand.
AI’s predictive power
The importance of AI in marketing is significant, as AI is no longer just a creative tool; it's becoming a market oracle, predicting what will sell even before products hit the shelves.
This predictive power is reshaping product design, marketing strategies, and even supply chains.
It’s a win-win situation for both consumers and sellers – enabling consumers to access the desired products, while helping sellers identify successful items and manage inventory more effectively.
A recent success story is L’Oréal, which uses AI to predict the skincare and makeup trends that will dominate the market. It then tailors its product lines accordingly.
Read more: AI, data, and the acceleration of drug discovery
What does the future hold for AI in marketing? I believe the future will feature a more efficient version of AI experiences across all industries and brands.
What we currently see with AI is primarily associated with certain brands, such as L’Oréal and Nike. However, in the future, AI will be integrated into various industries more broadly.
For instance, Spotify already utilises AI playlist models to predict which music tracks will become popular and influence artists’ creative direction.
This trend can also be observed in other industries, such as online retail, where websites provide personalised options based on your needs. As a result, we can expect more customised products tailored to individual preferences.
As such, AI will enable brands to offer hyper-personalised products using data from past purchases, browsing habits, and customer feedback.
Forecasting product demand
Nevertheless, not all data used for AI predictions must come from personal experiences or data. AI can also predict product demand across different markets and seasons, optimising inventory levels.
For example, Zara utilises AI for supply chain logistics to anticipate which fashion styles will be popular, ensuring a quick turnaround in production.
In the coming decade, we’re likely to see AI not only predicting trends, but also creating entire ecosystems of products and experiences based on subtle market signals.
The advancement of AI’s predictive capabilities brings significant power, but it also raises serious ethical, moral, and legal concerns that I find troubling.
If AI knows what customers want before they do, does this cross the line into manipulation rather than prediction? You might be wondering why this even matters.
It matters, as AI might exploit psychological triggers (for example, fear of missing out, impulse buying) to drive unnecessary purchases.
We must consider whether exploiting consumers’ subconscious behaviours for profit is unethical. In an era focused on greater sustainability and the circular economy, brands need to ensure they operate with transparency.
Ethical and legal concerns
This raises the question: Are customers making free choices if AI algorithms predict and influence their behaviours?
In addition to ethical concerns, there are also legal issues to consider.
How does AI make predictions based on individual experiences? AI’s capability to predict outcomes based on personal experiences comes from analysing extensive amounts of personal data, including browsing habits, purchase history, emotional triggers, and online behaviour patterns.
For example, L’Oréal uses AI-powered apps such as Skin Genius to analyse customers’ skin conditions through selfies. The AI recommends personalised skincare routines based on the analysed data.
However, this requires storing highly personal data (such as facial images), raising questions about data security and how this data is shared or monetised.
Brands must inform users about:
-
What data is being collected
-
How it will be used
-
Will it be shared with third parties?
Many privacy laws, such as the GDPR in Europe, require transparency and data subject consent as a legal obligation. However, the issue is that, in many cases, brands conceal data-sharing policies within lengthy terms and conditions, making it difficult for consumers to make informed choices.
The second legal issue relates to how AI models use their trained datasets, and whether the models have the authority to do so.
When generating predictions, AI may infringe on existing intellectual property (for example, designs, artwork, or ideas). For example, if an AI replicates a fashion design pattern predicted to trend, it might violate copyright laws.
This is where the question arises: How can major companies such as Adidas ensure its AI-based advertisements are free from copyright claims?
Have we considered what happens if predictive AI overpromises or misrepresents product features? What if the “floral-coloured” shoes showcased in an Adidas ad can never be produced as advertised?
Transparency is vital
Misleading AI predictions could violate consumer protection laws if customers are deceived about a product’s value or utility.
So, what should companies keep in mind before creating AI-driven advertisements? The answer is transparency, transparency, and transparency.
This is essential to address any ethical or legal claims relating to AI in marketing. Companies should disclose when AI is involved in making product predictions or ad campaigns, including the models and data used to create such advertisements.
As calls for AI regulations grow, I suggest implementing regular AI audits. These audits, conducted by regulatory or third-party organisations, can assess AI systems for bias, fairness, and transparency, ensuring AI is used in marketing legally and ethically.
Read more: Friend or foe? The problem with AI and greenwashing
Is it as simple as it sounds? Not at all – numerous questions remain unanswered.
For instance, why would a developer willingly share the AI models they’ve worked on? Isn’t it also a part of the creator’s copyright protection? Who should be informed about the legality and ethics of the data used to train the model? Who will conduct the AI audits – governments, regulatory bodies, or third parties (similar to the role of chartered accountants in the accounting field)?
These are essential questions, and like many researchers, I’m also seeking answers. The more we progress, the more we’ll discover.
Until then, let’s enjoy the exciting advancements of AI in 2025, especially in the field of AI in marketing.
