Five years after data scientist Frances Haugen brought forward approximately 20,000 pages of internal documents revealing the hidden mechanics of Facebook, her story has returned to the cultural spotlight with Aaron Sorkin’s dramatic film, The Social Reckoning. Starring Mikey Madison as Haugen, the movie synthesizes over 1,000 pages of non-fiction memoirs to explore both the technical anatomy of the leaks and the internal culture of Meta.

In a candid conversation on BBC News’ Top Comment Podcast, Haugen reflected on what has changed since her historic disclosures—and why, in many ways, Meta’s leadership has doubled down on its posture of defensiveness rather than fundamentally altering its incentives.

From Mark Zuckerberg’s isolated empire to multi-hundred-billion-dollar lawsuits and the urgent need for independent oversight in artificial intelligence, Haugen provided an inside look at the ongoing struggle to hold Big Tech accountable.

1. The Core Misconception: “Content Agnosticism” vs. Deliberate Inaction

At the heart of the disclosures Haugen made in 2021 was a fundamental question: Does Meta knowingly push harmful, divisive content to maximize profit?

Meta has long maintained that its algorithms are “content agnostic”—meaning they do not favor specific political topics or viewpoints, but simply display what users engage with most. However, Haugen argues that this defense falls apart when executives are presented with internal research proving how those algorithms function.

“If you get told that your algorithm has a bias towards encouraging content that is emotionally evocative—and you know the shortest path to a click is anger or divisiveness—and you choose not to fix it, at some point it becomes a knowing choice. They may not have set out to intentionally make things worse, but when told they had options to make things better, they chose not to take them.”

— Frances Haugen

The Global Moderation Deficit

This algorithmic bias had far-reaching consequences outside the Western world:

• Disproportionate Budget Allocation: While only 11% of Facebook’s global user base spoke English at the time of the leaks, over 50% of the company’s safety and integrity budget was allocated to English speakers.
• Unprotected Global Languages: Out of thousands of global languages, only a couple dozen had automated safety systems or human content moderation.
• Real-World Devastation: In countries like Myanmar and Ethiopia, where Facebook effectively was the internet due to zero-rating data agreements, bad actors and military regimes ran unchecked information operations that fueled real-world civil conflict and ethnic violence.

2. Inside the Mind and Compound of Mark Zuckerberg

Haugen offered a rare psychological assessment of Meta’s founder and CEO, Mark Zuckerberg. Unlike traditional corporate CEOs who are evaluated by independent boards and answer to voting shareholders, Zuckerberg holds the majority of Meta’s voting shares. Every major decision inside the company flows directly through him.

Having served as CEO since age 19, Zuckerberg has lived an existence largely insulated from standard feedback loops:

• Absence of Consequence: Psychological research shows that adults mature and develop moral agency by facing consequences—seeing the direct pain or impact their actions cause others. Because Zuckerberg cannot be fired or overruled by his board, he has lived without typical professional accountability.
• Physical & Social Isolation: Reporting from The Wall Street Journal and The New York Times reveals an increasingly insular lifestyle. Zuckerberg merged multiple properties in Palo Alto into a single compound, established a private one-room schoolhouse pod for his children, and constructed a massive, bunker-equipped estate in Hawaii.
• A Disconnect from Public Sentiment: Haugen pointed to Zuckerberg’s recent product announcements as evidence of this isolation. When launching a camera-less version of Meta’s smart glasses, rather than addressing widespread public anxieties about clandestine filming in intimate spaces, Zuckerberg brushed off privacy concerns on camera, stating the camera was omitted simply because it made the glasses cheaper to produce.

3. Retaliation and “The Dirty Secret” of Moderation

While Haugen’s disclosures were carefully structured through documented papers that could stand on their own legally, Meta took a far harsher approach toward other internal critics whose claims rested on personal testimony:

• Sarah Wynn Williams (Careless People): An early foreign policy advisor to Facebook who witnessed key decisions surrounding the Arab Spring, Williams published a memoir detailing her direct interactions with Zuckerberg and Sheryl Sandberg. Meta pursued aggressive legal penalties against her—accumulating hundreds of millions of dollars in potential fines—and imposed sweeping gag orders that even attempt to restrict her physical presence at public panels.
• Daniel (Kenyan Content Moderator): A moderator who led a landmark lawsuit on behalf of workers in Kenya, exposing severe mental health trauma, lack of psychological support, and exploitative contracting conditions.

Haugen described human content moderation as “one of the dirty secrets of the internet”—a hidden layer of low-wage workers in global processing centers forced to review graphic violence daily to keep public feeds clean.

4. The Regulatory Shift: From Dismissive Letters to $300B Lawsuits

Despite Meta’s internal resistance, public perception and regulatory pressure have shifted dramatically over the past five years:

2021 Baseline	Present Reality
Dismissive Responses: US Senate inquiries regarding youth mental health received non-binding boilerplate letters from Facebook executives.	Legally Mandated Limits: 47 US states have established consensus on digital childhood harms, leading to mandatory 2-hour daily Instagram caps for users under 16 and bans on push notifications during school hours and late nights.
Whistleblower Accusations: Allegations were largely treated in the media as the subjective opinion of an isolated engineer.	Jury-Backed Liabilities: A massive class-action lawsuit brought by school districts seeking over $200B to $300B in damages for security, tutoring, and mental health costs is moving through US courts, backed by direct jury verdicts against tech platforms.

5. From Social Media to Frontier AI: The Last Line of Defense

As the tech landscape transitions from social networks to generative AI models and synthetic agents, the lessons of the Facebook Files are being applied to frontier AI safety.

Haugen noted that tech whistleblowing has evolved significantly. While consumers can disassemble an iPhone at home to inspect its hardware, complex black-box AI models like ChatGPT or Claude cannot be inspected from the outside.

Because of this opacity, internal employees at AI labs have become the primary “last line of defense” before formal governance exists:

• Normalizing Dissent: Unlike a decade ago when tech was viewed as inherently harmless, tech workers now recognize their ethical obligation to raise alarms publicly when safety protocols are bypassed.
• Embedded Independent Auditors: Haugen endorsed proposals—recently supported by Anthropic CEO Dario Amodei—calling for independent, external auditors to be embedded inside frontier AI companies with employee-level data access to verify safety claims independently.
• Rehabilitating the Digital Childhood: With children as young as 7 to 9 occupying social platforms, and teenagers increasingly forming primary emotional relationships with synthetic AI personas, society faces an urgent need to rebuild spaces for in-person human connection.

Conclusion: Lessons from The Social Reckoning

Aaron Sorkin’s The Social Reckoning serves as more than a cinematic retelling of the 2021 leaks; it captures the lonely, high-stakes reality of standing up to one of the most powerful corporate monopolies in human history.

As Frances Haugen emphasized, meaningful accountability will not come from expecting tech leaders to undergo spontaneous moral awakenings. Instead, it requires changing the structural incentives, enforcing external transparency, and recognizing that public safety must always take precedence over algorithmic engagement.