A rogue Meta AI ad tool is at the center of a new controversy after outdoor retailer REI faced severe backlash for an Instagram campaign featuring a physically impossible bicycle. Users quickly noticed that the bike in the promotional image was equipped with two distinct sets of handlebars, prompting immediate mockery online. The incident highlights the growing friction between automated advertising platforms and brand safety.
The backlash gained traction over the weekend when a Reddit user posted the image with the caption, REI using AI slop now. So much for caring about the environment. The post quickly amassed hundreds of upvotes, forcing the retailer to address the bizarre creative choice. According to REI, the image was not an intentional design but the result of an automated generative AI feature.
Meta auto-enrolled us in an AI personalization tool that produced an inaccurate and inappropriate alteration of a vendor-provided image in some of our ads.
- Spokesperson, REI
The original image was intended to promote a bicycle from the brand Van Rysel. Van Rysel North America confirmed that the source photo was entirely authentic, originating from a legitimate photo shoot featuring professional cyclist Amity Rockwell. The vendor explicitly stated that they did not make any subsequent alterations to the image before it was processed by the advertising platform.
Following the incident, REI confirmed it has unenrolled from the Meta tool, apologizing for the confusion and emphasizing that product accuracy is critical to its vendor relationships. Meta declined to comment on the specific incident, but the company's terms for generative AI ad tools explicitly state that outputs may be "inaccurate, incomplete, misleading, offensive, and/or inappropriate," placing the burden of evaluation entirely on the advertisers.
This is not an isolated glitch. Multiple advertisers have previously reported that Meta's AI tools generated nonsensical ads for their products without explicit consent. Settings such as "test new creative features" and "automatic adjustments" have reportedly been toggled on by default for many accounts, forcing brands to actively opt out to protect their visual identity.
The Hidden Cost of Auto-Enrolled AI
Meta's aggressive push to automate ad creation is clearly prioritizing scale over brand safety. By defaulting to an "opt-in" model for generative AI features, platforms are forcing brands to actively police their own campaigns against algorithmic hallucinations. REI's two-handlebar disaster is a textbook example of how AI lacks the contextual awareness required for physical product accuracy.
While Meta claims these tools improve overall ad performance, the reputational damage of publishing "AI slop" can easily outweigh the benefits of automated A/B testing. Until generative AI can reliably distinguish between a creative background enhancement and a physically impossible product mutation, brands must treat auto-enrollment features as a direct liability.
