How Your Family Photos Are Now Fueling AI

The Evolving Dangers of Sharenting

The ethics of sharing photos online, especially those of children, have changed dramatically. Not long ago, posting back-to-school photos or holiday pictures was a common practice. However, a noticeable shift is occurring, with even the most prolific sharers on social media pulling back. This change is driven by the evolving business and politics of “sharenting,” a term that once humorously described oversharing parents but has since taken on a more serious tone. The advent of AI is set to make these considerations even more complex.

The term "sharenting" now carries a sharper edge, fueled by stories of teenagers asking their parents to remove childhood images posted without their consent. These young people often feel embarrassed or distressed by a permanent online record of their vulnerable moments, from meltdowns to being in diapers.

Former child influencers, whose lives were turned into monetized content, have been particularly outspoken. Cam Barrett, a TikTok influencer, testified at a hearing in Washington state, expressing fear of using her real name due to a “digital footprint I had no control over.” She detailed how her mother shared private information, including her first period, illnesses, and a car crash. This has spurred a political response, with France passing laws to protect child influencers' earnings and image rights, and several US states considering similar legislation. As a result, many parents are now more cautious, opting for private messaging apps or photos that obscure their children’s faces.

The Insatiable Appetite of AI for Your Data

A significant new development is underway. If a photo of you or your child is on the internet, it has likely been used to train AI models that power generative tools. The concern is no longer just about who might see your images, but also how they are being processed by machines.

Generative tools like OpenAI’s ChatGPT and Google’s Gemini are built on AI models that require almost unimaginable amounts of data for training—billions of texts, photos, and videos. Initially, tech companies scraped the public web for this data. Having exhausted that resource, they are now in a race to find new content sources to dominate the next era of the internet.

This hunger for data is so intense that Meta reportedly considered acquiring the publishing house Simon & Schuster for its content library. Ultimately, according to court filings, they allegedly turned to downloading millions of pirated books and research papers. This has led to lawsuits from creators, from journalists to illustrators, who are suing AI companies for using their work without permission, centering on the murky legal question of whether data scraping constitutes “fair use.”

Social Media Feeds Are the New AI Goldmine

It was perhaps inevitable that tech giants would turn to their own platforms for data. Google transcribed millions of YouTube videos to train its models, a move that also led to OpenAI being sued by creators. Meta is now exploring how to leverage the trillions of posts and photos uploaded to Facebook and Instagram over the decades. In 2024, Mark Zuckerberg told shareholders that Meta’s own data set is larger than Common Crawl, one of the biggest web data sets. He was referring to your public photos and posts.

This could extend even further. Reporting by TechCrunch and The Verge suggests that changes to Meta’s terms of service might allow the company to use unpublished photos from your phone’s camera roll for AI training. While Meta states this is currently an opt-in feature, the possibility remains open for the future.

Every Photo Is Now a Commodity

All of this means the politics surrounding our photos—shared or not—have become incredibly fraught. We've moved beyond worrying about malicious individuals stealing images to confronting the ethics of our children's lives feeding energy-intensive AI for purposes we can't yet fathom.

Ultimately, we must decide if we are comfortable with our private memories being used to boost the stock value of the world’s most powerful corporations. The monetization of childhood is no longer just for influencers; every photo you've ever taken is now a valuable commodity.

This article is based on the work of Liz Carolan, who writes about democracy and technology at thebriefing.ie.