the women we study: frances haugen

the architecture of conscience

There is a version of the whistleblower story that the culture has decided it prefers: the employee who reaches a breaking point, walks out in a moment of righteous anger, and calls a journalist. It is a satisfying narrative because it maps onto what we understand about moral courage — the instant when a person decides that they cannot participate in something wrong for one more day.

Frances Haugen did not do that. What she did was more deliberate, more methodical, and ultimately more consequential than a moment of conscience. She spent months inside Facebook building a case. She organized tens of thousands of internal documents around a specific argument. She hired a whistleblower attorney before she told anyone what she was planning. She contacted regulators, coordinated with journalists, and prepared testimony for multiple governments — all before she revealed her identity to the public. She was not acting on impulse. She was constructing something that could not be dismantled.

That distinction matters enormously, and it is the thing the narrative about her most often flattens.

Haugen grew up in Iowa, the daughter of two artists who raised her with a particular orientation toward honesty — not as an abstract value but as a practical one, something you owed to the people around you and to yourself. She studied computer engineering at Olin College and earned an MBA from Harvard Business School, spending years building algorithms at Google, Pinterest, and Yelp before she arrived at Facebook in 2019. She joined as a product manager on the civic integrity team, the group responsible for understanding how the platform was being weaponized to spread misinformation and undermine democratic processes. She was not a naive hire and she was not unfamiliar with how algorithmic systems work. She understood what they optimize for, how engagement metrics operate as a proxy for value, and what the gap between those two things produces at scale.

What she found at Facebook was not a company that didn't know what its products were doing. It was a company that knew, produced internal research documenting what it knew, and made a different calculation anyway.

The research she read was specific and it was damning. Facebook's own scientists had studied the relationship between Instagram use and teenage mental health and had produced findings that the company had not published, had not acted on in any meaningful way, and had actively worked to minimize when they became public. One in every eight teenage girls who had experienced suicidal thoughts believed Instagram had made those thoughts worse. Thirteen and a half percent of teen girls said Instagram worsened their suicidal ideation. Seventeen percent said the platform contributed to their eating disorders. The company's own researchers had described the dynamic as Instagram's perfect storm — an algorithmic system that identified a teenage girl's insecurities in her first minutes on the app and then fed her progressively more extreme content designed to keep her engaged, each piece optimized to produce the emotional response that would make her stay longer, regardless of what that response was doing to her.

Facebook saw this research. Facebook's leadership reviewed it. And Facebook concluded that the cost of meaningful intervention — in user growth, in engagement metrics, in revenue — was higher than the cost of the harm the research described.

Haugen read all of this. She read it as a product manager who understood how recommendation systems are built, what decisions go into their construction, and what it means when an engineer makes a specific technical choice to optimize for a specific outcome. She understood that the algorithm leading teenage girls to anorexia content was not a bug. It was a feature of a system designed to maximize the time a user spends on the platform, applied to a population that had not developed the regulatory capacity to resist it. She has said that she thought about cigarettes — about the decades during which the tobacco industry produced internal research documenting the harm of their products, suppressed that research, and continued selling to consumers who were not equipped to protect themselves. The parallel was not rhetorical, it was structural.

She made a decision, and then a plan.

The planning took months. She copied tens of thousands of internal documents, working carefully and specifically, organized around the argument she intended to build. She identified what would be most useful to legislators, to regulators, to journalists, and to courts. She thought about what the company's defense would be — that she was a disgruntled employee, that she lacked direct knowledge of the child safety team's work, that the research was more nuanced than she was representing — and she built her case around the evidence rather than her own testimony, so that the documents would stand independently of whatever the company said about her.

She hired a whistleblower attorney before she told anyone what she was doing. She contacted the Securities and Exchange Commission. She reached out to congressional staffers. She coordinated with the Wall Street Journal, which published a blockbuster investigative series in September 2021 that became known as the Facebook Files — one of the most significant pieces of tech journalism in the platform era. She revealed her identity on 60 Minutes in October 2021 and testified before the Senate Commerce Subcommittee the following morning. She subsequently testified before the parliaments of the United Kingdom, the European Union, and four other governments, in each case bringing the same organized evidentiary record and the same methodical argument.

The company's response was exactly what she had anticipated. They called her claims misleading. They noted she had not worked directly on child safety or Instagram. They argued that their platforms did more good than harm, that the research she was citing was more ambiguous than she represented, that she was a single disgruntled employee with incomplete knowledge of a complex organization. They were not entirely wrong that her knowledge was partial — no single person inside a company of Facebook's size has complete knowledge of its operations. But the documents were not partial. The documents were the company's own research, produced by its own scientists, reviewed by its own leadership, and suppressed by its own communications teams. She had not made claims about what Facebook knew. She had provided the evidence.

Facebook paused development of Instagram for kids within days of her testimony. The United Kingdom passed the Online Safety Act in 2023, one of the most sweeping pieces of platform regulation in the world, and the internal Facebook research Haugen had provided was cited in the parliamentary debate that preceded it. The United States Senate moved children's online safety legislation that had stalled for years, and while comprehensive federal regulation of social media platforms remains incomplete, the evidentiary record Haugen built has become the foundation that subsequent regulatory efforts have built on. None of these outcomes happened because of one person. But they happened faster, and with more specificity, because one person had taken the time to organize the evidence into something that governments and regulators could actually use.

She has said that she knew before she left Facebook that she would not work in the technology industry again at the level she had reached. That the retaliation would not be personal but structural — that no major platform would hire someone who had done what she did, that she was closing a door permanently and she knew it before she walked through it. She had a coveted position, a Harvard MBA, and a career trajectory that most people in her field spend years working toward. She understood what she was giving up. She left anyway, because she had read the research on what Instagram was doing to teenage girls and she could not unknow what it said.

The women we study in this series are the ones who built something out of what they believed. Frances Haugen is a more complicated case: she dismantled something, and the dismantling required just as much architecture as any of the things we tend to celebrate. She understood that moral clarity is only the beginning of the work. That conscience without infrastructure is noise, and that the parents whose daughters were being led into eating disorder content by an algorithm that knew exactly what it was doing needed more than someone willing to speak. They needed someone willing to organize — to take the internal research, build it into an evidentiary record, and deliver it to the people with the power to act on it in a form they could not dismiss.

She did that. Methodically, at personal cost, with full knowledge of what it would mean for her career and her life.

We study her because she understood that the children were more important than the career. And because she understood that knowing that was not enough — you had to build something with it.