Anastasia Braitsik stands at the forefront of digital strategy, navigating the intricate balance between data-driven performance and the ethical deployment of emerging technologies. With a career dedicated to SEO and content marketing, she has become a primary voice for brands seeking to maintain integrity in an era where the line between human and synthetic creation is increasingly blurred. Her work helps global enterprises implement transparency standards, ensuring that innovation does not come at the cost of consumer trust. In this discussion, we explore the nuances of the updated IAB framework, the evolving regulatory landscape across different jurisdictions, and the psychological impact of AI disclosure on modern audiences.
When differentiating between routine post-production and synthetic content that alters authenticity, how should brands determine which AI-driven modifications require a formal disclosure?
The decision-making process really hinges on the concept of material authenticity and whether the technology changes what a consumer reasonably believes to be real. Under the IAB’s Version 2 framework, we don’t advocate for labeling every single instance where AI touched a project, as that would lead to “label fatigue” and dilute the message. Instead, brands must look at whether the AI has created realistic synthetic content that could fundamentally mislead the viewer, such as digital twins of deceased individuals or living people in fabricated situations that aren’t standard endorsements. For instance, if you are using generative AI to draft internal copy or perform standard audio enhancement, those are considered routine workflows that do not require a public disclaimer. However, the moment you introduce a synthetic voice or an avatar that a consumer could mistake for a human representative, the moral and legal obligation to disclose kicks in. It’s about protecting the identity and representation within the creative asset rather than just checking a box for using a tool.
With research showing mixed reactions to AI in advertising, how do you interpret the fact that more than half of consumers want clear disclosure for fully generated content?
The data from the Sonata Insights research is quite telling because it highlights a deep-seated desire for honesty in the digital space. While many consumers view AI’s role in creative work positively—seeing it as a frontier for innovation—there is a significant portion of the audience that feels a sense of inauthenticity when they realize they’ve been watching or listening to something entirely synthetic. When more than half of your audience explicitly asks for labels on fully AI-generated imagery or video, ignoring that demand becomes a massive brand risk. We see this play out most vividly with conversational AI; if a customer realizes they are talking to a chatbot they thought was a person, the sense of betrayal is palpable. By using the standardized sparkle icon or clear text disclosures, brands can actually lean into this transparency, turning a potential “gotcha” moment into a demonstration of corporate responsibility.
As we navigate a complex global landscape where laws like California’s SB 942 and the EU AI Act are now firmly in effect, how should marketing operations handle the “classification problem” of tracking AI usage?
The growing patchwork of international regulations has turned AI disclosure into a significant logistical challenge for marketing operations. It is no longer enough to simply know that AI was involved in a campaign; you must track exactly what that AI did and whether it altered the authenticity of the asset. For example, Article 50 of the EU AI Act requires disclosure for deepfakes and covered synthetic content, but it doesn’t mandate a specific icon, whereas the U.S. industry often leans toward the IAB’s recommended sparkle icon. This means a single global campaign might need different disclosure layers depending on where it’s served, whether that’s New York, South Korea, or the European Union. Operations teams now need to maintain a detailed ledger of AI “touches”—recording if the tool was used for a cartoon avatar, which is exempt, or for a realistic synthetic voice, which likely requires a label. This level of granular tracking is the only way to avoid the legal pitfalls that have emerged since these mandates took effect two years ago.
Given that the IAB framework rejects the idea that every use of AI needs a label, how can brands maintain a high standard of creativity while ensuring they don’t fall into “labeling debt”?
The goal is to maintain a balanced approach where we respect the consumer’s right to know without cluttering the creative experience with unnecessary warnings. We distinguish between “behind-the-scenes” AI—like generic synthetic voices, background music, or stylized cartoon avatars—and AI that mimics reality. If a marketer uses generative AI to brainstorm copy or polish a background, the nature of that use is fundamentally different from creating a video of a person doing something that never actually happened. By focusing on whether the AI materially affects the representation of a person or situation, brands can avoid “labeling debt,” which occurs when you spend more time managing disclosures than creating content. It’s about staying true to the nature of the AI’s involvement; if the technology is just a digital paintbrush, keep the focus on the art, but if it’s a digital masquerade, you must let the audience see behind the mask.
What is your forecast for the future of AI transparency in marketing?
I expect we will move toward a more unified global standard where the “sparkle icon” becomes as ubiquitous and understood as the “verified” checkmark on social media. Over the next two years, we will likely see a shift where disclosure isn’t just seen as a legal hurdle, but as a stamp of high-quality, ethically produced media. Consumers will become increasingly savvy at identifying synthetic content, and brands that have been transparent from the start will hold a much higher trust equity than those that tried to hide their use of generative tools. Ultimately, the classification systems we are building today will become automated features within the martech stack, making it seamless for creators to flag synthetic elements at the moment of creation, ensuring that transparency is baked into the creative process rather than being an afterthought.
