The Future of Digital Advertising
When AI Claims Put Brands at Risk
The spread of artificial intelligence (AI) is rapidly transforming the digital advertising landscape. For trademark and brand professionals, AI-washing, dark patterns, and agentic AI are reshaping how brands communicate. Sarah Morgan finds out more.
The worlds of artificial intelligence (AI) and advertising law have collided. The rise of AI-generated content means that the digital advertising landscape looks very different today than it did even five years ago, according to Jayde Wood, partner at Gowling WLG, LLP (Canada).
She explains: “To put that in perspective, ChatGPT had not even been released five years ago, and now AI has moved from a behind-the-scenes optimization tool to a front-facing feature that brands actively promote, and that shift has brought three big challenges to the fore.”
The first is the rise in AI-washing, says Ms. Wood, where companies make unsubstantiated claims about AI capabilities in their products. Much like greenwashing, jurisdictions around the world are now requiring advertising claims to be substantiated.
Ms. Wood will moderate a session exploring the intersection between AI and digital advertising in Digital Advertising Under the Microscope: AI Claims, Dark Patterns, and Global Legal Trends (Sunday, May 3, 10:00 am–11:00 am).
Ms. Wood will be joined by panelist Enrico Bonadio, Professor of Law at St George’s University of London (UK), who is struck by “how consistently jurisdictions are converging on stricter scrutiny of AI generated advertisements, influencers, green claims, and children’s marketing, despite very different legal traditions and enforcement cultures worldwide.”
For example, fellow panelist Tiffany Shimada, partner at Dorsey & Whitney, LLP (USA), points out that the U.S. Securities and Exchange Commission has “taken a proactive stance against AI-washing and has emphasized the need for full and fair disclosure of AI-related risks, limitations, and conflicts of interest.”
Courts have increasingly recognized the risks posed by AI-washing, particularly in the context of fabricated or misleading AI-generated content, including case-law hallucinations, she explains.
“In response, courts have imposed sanctions and disciplinary action against attorneys who submit false and misleading AI-generated information,” adds Ms. Shimada.
Ms. Wood summarizes the current stance for brand owners: “The bottom line for brands is to treat AI claims like health claims: verify them before making them and check whether affirmative disclosures are required.”
"Courts have imposed sanctions and disciplinary action against attorneys who submit false and misleading AI-generated information."
- Tiffany Shimada | Dorsey & Whitney, LLP (USA)
Dark Patterns and Agentic AI
Second, real brand liability is being created by the spread of AI-driven dark patterns, which are manipulative design practices that use AI to steer, pressure, or nudge users into choices they would not otherwise make, often without their awareness, explains Ms. Wood.
“AI optimizes manipulative ‘dark pattern’ interfaces at scale by personalizing friction, nudging, and choice architecture. These practices not only invite scrutiny but also erode consumer trust and long-term brand equity when they are exposed,” notes Mr. Bonadio.
Ms. Shimada adds: “As noted in case law, reputational harm and loss of business opportunities can constitute irreparable harm, particularly when it results in the loss of client relationships and customer goodwill."
The third challenge, says Ms. Wood, is the emergence of agentic AI. This type of AI can autonomously plan, decide, and act to achieve a goal, rather than simply responding to prompts or generating content on request.
She notes that this is beginning to reshape how consumers interact with brands.
Looking to the near future, Mr. Bonadio says: “Agentic AI will autonomously negotiate, monitor, and enforce brand and platform terms, turbocharge clearance and watching, but also generate registrability, authorship, and accountability puzzles as bots increasingly ‘act’ in commerce under brand identities.”
"The bottom line for brands is to treat AI claims like health claims: verify them before making them and check whether affirmative disclosures are required."
- Jayde Wood | Gowling WLG, LLP (Canada)
A leading figure in the debate around AI inventorship and patents, Ryan Abbott, is also on the panel. He is Professor of Law and Health Sciences at the University of Surrey Law School (UK) and agrees that the rise of agentic AI will throw up even more legal queries.
He says: “When an AI agent selects products based on data and specifications rather than brand perception, established doctrines around consumer confusion and functionality start to wobble. That is a fascinating challenge. Practitioners should be asking how their clients’ brands perform when the ‘consumer’ is a machine, because portfolios may need to work for algorithms, not just people.”
He adds: “AI is expected to both generate a huge amount of content that brands will want to protect with trademarks and other types of intellectual property, and AI will also become increasingly capable of filing and prosecuting—or assisting with filing and prosecuting—trademark applications and related sorts of filings.”
Mr. Abbott is lead counsel for the DABUS litigation, a series of high-profile test cases around the world asking a single core question: Can an AI system be named as an inventor on a patent?
He says: “The DABUS decisions are about subsistence of IP rights in AI-generated and AI-assisted content, so this has significant relevance for branding practitioners, and the relevance becomes greater with more reliance on AI in branding.”
He adds: “Practitioners need to be aware of different rules in different jurisdictions related to the ability to own, license, and monetize AI content whether under trademark, copyright, or patent laws.”
Ms. Wood notes that for trademark and brand practitioners, the DABUS litigation raises several important questions around the ownership gap. For one, if AI can’t be an inventor or author, can assets generated by AI be registered and enforced? As for the infringement risk, when AI tools trained on existing marks produce something confusingly similar, who is liable?
According to Ms. Wood, “These are not hypothetical concerns; they are real, live issues that will shape portfolio strategy for years to come.”
She adds: “The DABUS cases have shown that courts everywhere are asking hard questions about what AI actually does. Regulators are now bringing that same skepticism to advertising, treating ‘AI-powered’ claims not as puffery but as statements that need substantiation.”
"Practitioners should be asking how their clients’ brands perform when the ‘consumer’ is a machine, because portfolios may need to work for algorithms, not just people."
- Ryan Abbott | University of Surrey Law School (UK)
