Urgent.News

What's breaking now, across thousands of outlets.

AI

Is that advertisement AI-generated? These 6 signs offer clues.

Real advertising examples reveal six clues that an ad might be AI-generated, from distorted products and garbled text to inconsistent characters.

Is that advertisement AI-generated? These 6 signs offer clues.

The rise of AI-generated advertisements is becoming increasingly common, appearing in everything from local sandwich-board ads to high-budget campaigns and political races. However, distinguishing between authentic and AI-generated images is becoming more challenging. A September study of over 500 consumers revealed that only 17 percent accurately identified an AI-generated ad among a lineup of ten, despite half initially feeling confident in their ability to spot it. Several subtle indicators can reveal whether an ad is AI-generated.

Firstly, the product may appear plausible overall, but its individual components may become inconsistent. For instance, an advertisement featuring a bicycle with an extra set of handlebars growing out of its saddle may have a believable setting and model, but the product itself is implausible and not found in the retailer's assortment. The company may attempt to explain away the issue, stating that Meta's AI personalization tool altered the vendor-provided image.

Secondly, characters or objects may appear differently across shots. This inconsistency can occur in the main character or vehicle. In Coca-Cola's 2025 AI-generated "Holidays Are Coming" commercial, the vehicles exhibited changing designs and mechanically implausible configurations. Similar inconsistencies were observed in Toys "R" Us's 2024 film, in which the appearance of the child portraying the company's founder varied throughout the video.

Thirdly, movement may not align with the object or speech. This issue was evident in Coca-Cola's 2024 AI holiday advertisement, where trucks moved along roads without their wheels rotating. Likewise, McDonald's Netherlands faced criticism for an AI-generated Christmas commercial featuring distorted faces and bodies. In cases of synthetic individuals, paying attention to discrepancies between the face and voice can help detect AI involvement, as poor synchronization of lip movements and flickering eyes around the mouth can indicate the use of deepfake technology.

Lastly, lettering and smaller details often fall apart in AI-generated ads. While the overall image may look convincing, elements such as product labels, objects on shelves, scenery, and body anatomy may appear nonsensical upon closer inspection. This is evident in the promotional images for Glasgow's "Willy's Chocolate Experience," where the AI-generated ads promoting the event initially promised nonsensical details.

Written by urgent.news from Mashable's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at mashable.com →

More in AI

You say you don't want AI. Google's school fight shows what that really means.

Google's AI dispute with schools exposes how much of the internet already depends on AI, and what rejecting it would mean.

  • Google CEO Sundar Pichai plans to restrict AI access for schools rejecting the technology.
  • Move reveals challenge of teaching students to critically evaluate AI-generated content.
  • Rejecting AI may hinder development of essential research skills and fact-checking abilities.

More from Friday 6 November →