Your Sadness Is Someone Else's Ad Timing

Opening

Dear reader, a teenage girl takes a selfie, doesn’t like it, and deletes it. A few seconds later, a beauty product ad appears on her screen. Coincidence?

In April 2025, testimony before a US Senate hearing suggested it might not be. Sarah Wynn-Williams, Meta’s former Global Director of Public Policy, told senators that Meta had identified the emotional states of teenagers aged 13 to 17 and passed that information along to advertisers — capturing the exact moments they felt “worthless,” “helpless,” or “like a failure.”

But there’s an uncomfortable question buried in this story. If the targets had been adults instead, would this not be a problem at all?

What Happened — The 2017 Leak, The 2025 Testimony

The roots of this story go back to 2017, when the Australian newspaper The Australian obtained and reported on internal Facebook documents. The documents showed that Facebook had been monitoring teenage users in Australia and New Zealand, identifying moments when they needed “a confidence boost,” and pitching advertisers on access to that data.

At the time, Facebook officially responded that it did not offer advertising targeting tools based on emotional state. The company said the internal analysis was meant to help marketers understand their audience, that it had never actually been used for ad targeting, and that it was based on anonymized, aggregated data.

But on April 9, 2025, Wynn-Williams’s Senate testimony called that explanation into question. Answering a question from Senator Blackburn, she stated that Meta had identified moments when teenagers aged 13 to 17 felt depressed or self-critical, and had shared that information with advertisers. When a teenage girl deleted a selfie, she’d be shown a beauty product ad; when she showed anxiety about her body shape, a diet product ad would appear.

More notable still is additional material released in September 2025 by Senator Grassley, chairman of the Senate Judiciary Committee. A 2014 internal presentation his office obtained, titled “Global Youth Study,” showed Facebook analyzing the emotional behavior patterns of 1,000 respondents aged 13 to 24 — and explicitly listing the intended audience for the findings as “influencers, agencies, brands, sales, other industries, clients.” This directly contradicts the 2017 claim that the data was “never used for ad targeting.”

To date, Meta hasn’t disputed the accuracy of these records themselves. It has, however, maintained a blanket denial that Wynn-Williams’s testimony as a whole is “divorced from reality and riddled with false claims.”

Is This Only a Problem “Because They’re Minors”?

The most intuitive reaction to this case is outrage — “how could they do this to children?” And that outrage rests on solid ground.

First, there’s the matter of cognitive development. The American Psychological Association (APA) has classified advertising directed at children under 8 as “inherently unfair.” That’s because children don’t yet recognize the persuasive intent behind advertising at all. “Advertising literacy”¹ typically doesn’t reach adult levels until around age 18, and the prefrontal cortex² doesn’t fully mature until around age 25.

Second, a legal framework already exists. In the US, COPPA (the Children’s Online Privacy Protection Act) requires parental consent to collect personal information from children under 13. States including Utah, Arkansas, and Texas have passed laws banning targeted advertising to minors under 18 outright, and KOSA (the Kids Online Safety Act)³ is moving forward with bipartisan support.

Third, research shows that the degree of emotional vulnerability differs. Adolescence is when a sense of identity is being formed. If a message saying “you’re not good enough” lands precisely in a moment of self-doubt during this period, its effect can be qualitatively different from an adult seeing the same ad.

So far, this much is clear: minors need special protection, and if Meta’s conduct is true, it’s a serious ethical and legal problem. But the story doesn’t end here.

Is It Fine If They’re Adults? — An Uncomfortable Extension

Among the internal chat logs released at the hearing is one exchange: a Facebook policy director asks a colleague, “Is it true we’re studying the emotional state of young mothers?” The colleague replies: “Yes.” Then, half-joking, adds: “Should I ask our morally bankrupt colleagues if there’s more research like this?”

The key point here: these “young mothers” are not minors. According to Wynn-Williams’s book Careless People, Facebook was even mapping emotional indices by race — for example, a “Hispanic and African American Feeling Fantastic Over-index.” This, too, had nothing to do with minors.

In fact, behavioral manipulation through emotional data is something Facebook has been researching for a long time. In 2014, Facebook ran an experiment manipulating the News Feeds of roughly 690,000 users. It demonstrated that reducing positive content led users to write more negative posts, while reducing negative content led them to write more positive ones. The paper was published in PNAS, but it drew fierce ethical criticism once it emerged that it had been conducted without participant consent. And the subjects of that experiment, too, were mostly adults.

This raises a structural question: have we grown so accustomed to the “protect the children” frame that we simply accept, as just “business,” the territory where that frame doesn’t apply — namely, emotional targeting of adults?

The Surveillance Capitalism Lens

Viewed through the “Surveillance Capitalism”⁴ frame that Harvard Business School’s Shoshana Zuboff put forward in 2019, the essence of this problem comes into sharper focus.

Zuboff’s core argument runs like this: digital platform companies treat user experience as free raw material, convert it into “behavioral surplus,” and sell it to advertisers. And as competition intensifies, this evolves beyond merely predicting behavior toward actively manipulating it.

Meta’s emotional targeting fits this frame precisely. The “worthlessness” or “anxiety” a user feels isn’t just a psychological state — it’s raw material that sharpens the accuracy of behavioral prediction. People make more impulse purchases when they’re depressed. When self-esteem is low, they’re more susceptible to messages like “buying this will make you feel better.” This mechanism doesn’t operate on teenagers alone.

This is why the “protect minors” frame matters, but isn’t enough on its own. It tries to divide the issue along a single line — age: under 18 needs protection, 18 and over is an autonomous adult capable of judgment. But emotional vulnerability doesn’t split neatly along age lines.

Is a young mother going through postpartum depression an “adult” capable of autonomous judgment? What about a 40-something office worker whose self-esteem has collapsed after losing their job? Or someone psychologically shattered after being a crime victim? Does tracking these people’s emotional states in real time and slotting ads into their most vulnerable moments become “legitimate business” simply because the target is over 18?

Oz’s Lens

In marketing, timing is everything. Showing a product to a customer at the exact moment they feel a need — that’s marketing 101. So what Meta did, looked at coldly, was push targeting precision to its absolute limit. “People buy more when they’re depressed” isn’t some new discovery in the ad industry. And in practice, Meta’s ad products are remarkably effective — from the advertiser’s point of view, anyway. Personally, if I had to choose between spending money on Google Ads or Meta Business Suite, I’d put it into Meta. It performs better on reach and conversion.

But what I find genuinely troubling is the asymmetry. Traditional marketing uses emotion too — Christmas ads stir nostalgia, insurance ads poke at anxiety. But those are messages aimed at an undifferentiated mass audience. The advertiser has no idea whether I, specifically, am feeling depressed right now. What Meta did operates on an entirely different level. It identifies the specific emotion a specific person feels at a specific moment, and targets that exact moment. This isn’t “well-timed advertising” — it’s the commodification of emotional states. And there was no age limit on who could be commodified. And Meta isn’t the only one doing this — it’s just getting this much attention because it’s the one facing lawsuits.

I agree 100% with the call to “protect minors.” But if we stop at that frame, we miss the point. The real question isn’t “shouldn’t we not do this to children?” — it’s “should we be allowed to do this to anyone at all?” It’s something I say often: your cognitive ability and judgment don’t suddenly upgrade the moment you turn 20, right? Why do we assume that becoming an adult means you should suddenly be exposed to every risk and responsibility?

Actually, Wynn-Williams’s testimony that Silicon Valley executives forbid their own children from using their companies’ products sums up this structure best. They know exactly how the system treats its users. And they make sure their own families, at least, are kept out of it.

Closing

The allegations around Meta’s emotional targeting hold a structural problem that goes beyond “let’s protect the kids.” The mechanism of capturing emotion in real time and using it for ad timing works the same way whether the target is 13 or 30. Yet the current legal and ethical framework addresses this issue through a single criterion: age.

If someone asked, “did you ever consent to the commodification of your emotions?” — most people, adult or minor, would answer “no.” And yet we’re already inside that system. No one agreed to it, but it’s being used with remarkable precision. Do you consent to this?

References & Further Reading

• Sarah Perez, “Meta whistleblower Sarah Wynn-Williams says company targeted ads at teens based on their ‘emotional state’”, TechCrunch, 2025.4.9. : The 2024 article that served as the starting point for today’s newsletter.
• Senate Judiciary Committee, “Grassley, Blackburn, Hawley Demand Answers on Meta’s Emotional Targeting of Children”, 2025.9.2. : Senate Judiciary Committee material that released additional evidence, including the 2014 ‘Global Youth Study.’
• Grassley to Meta, “Targeting-Teens” Letter, 2025.4.16. A letter containing primary source material, including text messages between Meta employees.
• “Is Facebook Targeting Ads at Sad Teens?”, MIT Technology Review, 2017.5.1. : An article covering the 2017 Australian leaked documents — the origin point of this whole issue.
• Kramer, A., Guillory, J., & Hancock, J., “Experimental evidence of massive-scale emotional contagion through social networks”, PNAS, Vol.111 No.24, 2014. : The paper on the emotional contagion experiment involving roughly 690,000 users. It sparked major controversy for being conducted without consent.
• Shoshana Zuboff, The Age of Surveillance Capitalism, PublicAffairs, 2019. : An essential book for understanding the structure of surveillance capitalism. Part Three, on “behavioral surplus” and “behavior modification,” connects directly to today’s topic. It has also been translated and published in Korean.
• Sarah Wynn-Williams, Careless People: A Cautionary Tale of Power, Greed and Lost Idealism, 2025. : The insider memoir that formed the basis for this hearing. It hasn’t been published in Korea, but you can get it as an ebook or international import via Amazon.
• TechPolicy.Press, “Transcript: Former Exec Sarah Wynn-Williams Testifies on Facebook’s Courtship of China”, 2025.4.10. : The full hearing transcript. The Q&A on emotional targeting is in the latter half.

The author, Kwangseob Ahn, is a professor of business administration at Sejong University and lead consultant at OBF (Oswarld Boutique Consulting Firm). He teaches statistics and data analysis — business data management and business analytics — while leading GTM and AI strategy consulting in the field, designing the seam between technology and business. He has published academic research on a memory architecture for AI dialogue systems (HEMA) and runs Daily Arxiv, a daily curation of global AI papers. He holds a master’s from Korea University’s Graduate School of Technology Management and a KMBA. He is the author of Homo Brainless: The People Who Outsource Their Thinking.

Footnotes

1. Advertising Literacy: the ability to recognize the persuasive intent behind advertising and evaluate it critically. Simply put, it’s the ability to notice, “ah, this is trying to sell me something.” This ability is still developing during childhood and adolescence. ↩

2. Prefrontal Cortex: the region of the brain responsible for decision-making, impulse control, and predicting long-term consequences. It typically doesn’t fully mature until around age 25 — which is why teenagers are more prone to impulsive judgment. ↩

3. KOSA (Kids Online Safety Act): a child online safety law under discussion in the US. It would require platform companies to design out content that promotes suicide, eating disorders, substance abuse, and similar harms to minors, starting at the design stage. ↩

4. Surveillance Capitalism: a concept put forward by Harvard professor Shoshana Zuboff. It refers to an economic structure in which digital platforms collect users’ behavioral data as free raw material and process it into behavioral-prediction products sold to advertisers. The key point is that it evolves beyond mere “prediction” into “behavior modification.” ↩