People Are Pushing Back Against AI They Never Asked For, and It Is Working

Key points

• A recent Gallup poll found that almost half of Americans aged 18 to 29 believe generative AI does more harm than good.
• Meta disabled an Instagram feature that let anyone create AI deepfakes of other users after three days of public outcry and viral criticism.
• Google pulled a tool that let people use AI to alter satellite images on Google Earth almost immediately after launch, following widespread backlash.
• LinkedIn, Snapchat and Substack have each added features to limit or flag AI-generated content on their platforms.
• Protesters across the political spectrum are organising against the data centres, the large warehouse-sized buildings full of computers that power AI tools, over concerns about environmental and economic harm.

Something shifted this year. People are angry about AI, and companies are rolling back features because of it.

A recent Gallup poll found that growing familiarity with generative AI, the technology behind text, image and video tools like ChatGPT, does not make people like it more. It does the opposite. For Americans aged 18 to 29, nearly half now say the technology does more harm than good.

Why are people so frustrated?

For most people, the core complaint is consent: nobody asked for this. Training data was scraped from the web without permission. AI-generated answers appeared at the top of search results without warning. Features showed up inside familiar apps overnight.

"Consent is really important," Meredith Broussard, a data journalism professor at New York University and author of Artificial Unintelligence, told Wired. "Tech companies have never really been sensitive to issues of consent."

Nick Seaver, an associate professor of anthropology at Tufts University, puts it plainly. Companies "just roll them out and see what sticks," he says, "because nobody really knows what this is for."

The frustration is not limited to tech-savvy users. Local businesses have faced criticism for promoting events with AI-generated posters. Major brands including McDonald's and Coca-Cola drew sharp online backlash for using AI-generated visuals in advertising.

Is the backlash actually changing anything?

Yes, in concrete ways. Meta switched off a tool that let anyone generate AI deepfakes, fake images or videos of real people, of Instagram users after just three days, following viral videos with millions of views condemning the feature. Google pulled a similar AI image-editing tool for Google Earth almost immediately after launching it.

Platforms are also adjusting their rules. LinkedIn added a reporting button for AI-generated content. Snapchat announced that fully AI-generated videos no longer qualify for its main discovery feed. Substack added an AI detection tool to monitor what writers publish.

Beyond the internet, protests against data centres are growing. Emily Bender, co-author of The AI Con, notes that data centres draw opposition from communities across the political spectrum because they concentrate both environmental costs and economic disruption in one place.

What can ordinary people actually do?

"Organize using traditional methods as well as digital methods," says Broussard. "Remind people that public pressure works."

That is not empty encouragement. The rollbacks above happened because enough people complained loudly and visibly. Reporting unwanted AI features through official feedback channels, supporting platforms that offer genuine opt-outs, and joining local groups opposing data centre construction are all concrete steps that have demonstrably moved companies to act.

Collective action is slow. But this year's results suggest it moves faster than most people expected.