The Backlash Against AI Slop Is Making a Difference

I recently found myself navigating through my Instagram settings, attempting to disable Meta’s outdated feature that permitted anyone to create AI deepfakes of me without permission. Frustration welled up as I scrolled through the convoluted opt-out process.

Instead of hurling my phone, I exclaimed to my partner that we should abandon San Francisco and pursue a life as off-grid sheep farmers. Something tangible. In that moment, anything felt preferable to this ongoing nightmare induced by generative AI.

Many Americans share similar feelings about the increasing presence of generative AI in their everyday lives. “The AI revolution has occurred, and everyone despises it,” remarks Meredith Broussard, a professor of data journalism at New York University and author of Artificial Unintelligence: How Computers Misunderstand the World. According to a recent Gallup poll, a heightened familiarity with generative AI among Americans coincides with negative perceptions of the technology, particularly among those aged 18 to 29, nearly half of whom view generative AI as causing more harm than good.

Though this backlash has been simmering for a few years, more companies seem to be addressing these issues directly. LinkedIn introduced a “seems like AI slop” button for reporting AI-generated content, even while continuing to host various generative AI tools. Snapchat declared that fully AI-generated videos would no longer qualify for its discovery feed. Substack has recently implemented an AI detection tool to monitor writers.

The root of the backlash is straightforward. Many Americans feel they never agreed to any of this. I was unaware that my online activities would eventually be harvested en masse for AI model training. I didn’t authorize Instagram to enable deepfakes on my account. I didn’t wish for Google to prioritize AI-generated responses in search results. One day, I realized that generative AI features had infiltrated every software I depend on without my approval.

Whether it’s AI models mining your data or disruptive data centers emerging in your neighborhood, the AI revolution feels like a phenomenon happening to us, leaving little room for us to be active participants.

“Consent is really essential,” Broussard tells WIRED. “We can relate this to the exploitation and abuse faced by women, queer individuals, people of color, and minorities online. Tech companies have never truly been mindful of consent issues.”

Public backlash seems to be effective, at times, in eliminating unwanted AI features. “They simply deploy these features and see what resonates, because nobody is certain what their purpose is,” explains Nick Seaver, an associate professor of anthropology at Tufts University focusing on technology and society. Following 404 Media’s reporting, Google swiftly revoked the ability for people to use generative AI to modify satellite images on Google Earth shortly after launching it. Additionally, Meta disabled the aforementioned deepfaking tool for Instagram users just three days after widespread public criticism and viral videos garnering millions of views.

There’s also an increasing rejection of generative AI in marketing campaigns. Many people online voiced their disapproval when major brands like McDonald’s and Coca-Cola incorporated generative AI visuals into their advertisements. Local businesses, too, faced backlash for promoting events with AI-generated posters.

Beyond online protests, more individuals are starting to oppose the construction of data centers that support these AI tools. “There’s considerable pushback surrounding data centers,” states Emily Bender, coauthor of The AI Con. “Data centers are crucial because they represent focal points of environmental and economic harm.” Recently, Americans across the political spectrum have united in protests against these data centers.