AI backlash forces Meta, Google and LinkedIn to roll back features
A Wired writer opens with a personal anecdote: struggling to opt out of Meta's now-defunct feature that let anyone generate AI deepfakes of Instagram users without consent. The piece then argues that this kind of anger is widespread and is starting to change what tech companies actually ship. Meredith Broussard, a data journalism professor at New York University and author of Artificial Unintelligence: How Computers Misunderstand the World, is quoted saying "the AI revolution has happened, and everybody hates it." The piece cites a recent Gallup poll showing that as Americans grow more familiar with generative AI, their attitudes toward it get more negative, a trend that is sharpest among 18 to 29 year olds, almost half of whom see generative AI as doing more harm than good.
The article lists several concrete rollbacks. LinkedIn added a "seems like AI slop" button letting users flag AI-generated content, even though the platform still hosts its own generative AI tools. Snapchat announced that fully AI-generated videos will no longer be eligible for its discovery feed. Substack added an AI detection tool aimed at policing writers. Google pulled the ability to use generative AI to edit satellite images on Google Earth almost immediately after launching it, following reporting from 404 Media. Meta shut down its Instagram deepfake tool after three days of public outcry and viral videos, some with millions of views, criticizing the feature.
Broussard frames the core objection as consent: people never agreed to have their data scraped for AI training, to have deepfake tools enabled on their accounts, or to have AI answers pushed to the top of search results. "Consent is really important," she tells Wired, tying it to a broader pattern of tech companies being insensitive to consent issues affecting women, queer people, people of color and other minorities. Nick Seaver, an associate professor of anthropology at Tufts University, describes the pattern behind these features as companies shipping first and figuring out demand later: "They just roll them out and see what sticks, because nobody really knows what this is for."
The piece also points to backlash against generative AI in advertising, with brands like McDonald's and Coca-Cola drawing criticism for AI-generated visuals in ad spots, and even local shops facing pushback for AI-generated event posters. Beyond online complaints, the article describes growing protest against AI data centers. Emily Bender, coauthor of The AI Con, says "there's a lot of pushback coalescing around data centers" because they are "focal points of environmental and economic damage," and the piece notes that Americans across the political spectrum have united around these protests.
On workplace use, an unnamed Block employee, speaking to the writer during layoffs at Jack Dorsey's fintech company earlier this year, criticized mandatory AI tool use: "top-down mandates to use large language models are crazy. If the tool were good, we'd all just use it." The article closes on Broussard's case that organizing works: she encourages people to combine traditional and digital organizing methods, arguing that "public pressure works," and points to the recent feature rollbacks as evidence.
Key facts
- A recent Gallup poll found that as Americans grow more familiar with generative AI, their views of it get more negative; almost half of 18 to 29 year olds see it as doing more harm than good.
- LinkedIn added a 'seems like AI slop' reporting button, Snapchat excluded fully AI-generated videos from its discovery feed, and Substack added an AI detection tool for writers.
- Google rolled back Google Earth's AI satellite-image editing almost immediately after launch, following reporting from 404 Media; Meta shut down its Instagram AI deepfake tool after three days of public outcry.
- Protests against AI data centers are spreading; Emily Bender describes the pushback as coalescing around data centers as focal points of environmental and economic damage, and the piece notes that Americans across the political spectrum are uniting around these protests.
- An unnamed Block employee, speaking during layoffs at the company earlier this year, criticized mandatory internal use of large language models as a 'top-down mandate.'
Why it matters
The piece argues that public anger at generative AI, long dismissed as background noise, is now producing measurable outcomes: multiple platforms have reversed or restricted AI features within days of visible backlash. Meredith Broussard frames the underlying complaint as a lack of consent, people never agreed to have their data used for training, to have deepfake tools switched on for them, or to have AI answers inserted into search results, and connects that pattern to a broader history of tech companies disregarding consent for marginalized groups.
Who it affects
The rollbacks named span LinkedIn, Snapchat, Substack, Google (Earth), and Meta (Instagram). The article also cites brands such as McDonald's and Coca-Cola facing criticism for AI-generated ad visuals, local shops facing backlash for AI-generated event posters, communities protesting the construction of AI data centers near them, and tech workers, including one Block employee, who object to mandatory AI tool use at work.
How to use it
The article's practical takeaway comes from Broussard: organized public pressure, combining traditional and digital organizing methods, has already moved large companies to reverse specific AI features within days, and she encourages readers to keep applying it rather than assume backlash is futile against well-funded tech firms.
How solid is it
The piece is Wired reporting built on named, credentialed sources: Meredith Broussard (NYU professor and author), Nick Seaver (Tufts anthropology professor), and Emily Bender (coauthor of The AI Con), plus one anonymous Block employee. It cites a Gallup poll and 404 Media's reporting on Google Earth, but the article itself does not give a date, sample size, or margin of error for the poll, and the piece is framed as analysis and opinion rather than a data-driven study.
Risks and caveats
Several details are left unverified in the source: no timeframe is given for when LinkedIn, Snapchat, or Substack introduced their respective features, the Block employee is not named, and it is not stated whether Meta's Instagram opt-out process, which the writer describes as frustrating to navigate, has since been fixed. The article itself carries no byline in the stored text.
“The AI revolution has happened, and everybody hates it.”
— Meredith Broussard, data journalism professor at New York University