EU's deepfake definition gaps spark retail AI debate
The European Union’s AI Act is facing a fresh challenge—not from deepfakes themselves, but from how they are defined. Retail leaders including Amazon, H&M, and IKEA are pushing for AI-generated advertisements to be exempt from transparency requirements, arguing that a lifelike sofa image in a marketing campaign isn’t a deepfake—even if it looks indistinguishable from reality. Meanwhile, Zalando reports that 90% of its marketing content is already AI-generated, exposing the widening gap between policy and practice.
When AI marketing isn’t “deep” enough for regulators
Eurocommerce, the trade association representing major retailers, is calling for a narrow interpretation of deepfakes in the EU AI Act. They contend that synthetic media used for product promotion should not fall under rules aimed at manipulative content, such as political disinformation. Yet this stance risks blurring the line between innocuous commercial creativity and deceptive media. As AI tools generate ever more convincing visuals—from virtual clothing models to photorealistic room settings—the distinction becomes harder to maintain.
Retailers adapt, but the rules lag behind
Zalando’s revelation that nearly all its marketing content is AI-generated highlights the rapid adoption of synthetic media in retail. While companies tout efficiency and customization, critics warn that unchecked AI-generated ads could mislead consumers about product authenticity or availability. The current debate centers on whether transparency rules should apply only to content designed to deceive—or to all AI-generated material that could be mistaken for reality.
For now, the EU faces a choice: tighten definitions to cover emerging use cases, or risk real-world consequences where shoppers can’t tell product images from digital fabrications.
Source: The Decoder. AI-assisted editorial synthesis — TechnoExpress.