LSE - Small Logo
LSE - Small Logo

Gerry Brown

Randall S Peterson

January 14th, 2025

The free speech you won’t find inside Meta

0 comments | 6 shares

Estimated reading time: 10 minutes

Gerry Brown

Randall S Peterson

January 14th, 2025

The free speech you won’t find inside Meta

0 comments | 6 shares

Estimated reading time: 10 minutes

Meta doesn’t give employees or board members the same free-speech leeway it now extends to its users. Gerry Brown and Randall S Peterson write that, internally, there is structurally no opportunity for departure from whatever content Mark Zuckerberg decides should go without community notes, fact checkers and freely expressed dissenting voices. They say the company’s board suffers from various dysfunctions, including a lack of independence.


On the Facebook and Instagram platforms owned by Meta, henceforth free expression will be restored and institutional censorship mostly disallowed. Mark Zuckerberg released a strident, often impassioned strategic message about these new-era freer-speech Meta policies now applying. In essence, content moderation will be significantly reduced in order to return the company to its newly discovered roots of giving people a voice.

On the surface, this could sound both laudable and benign, though it ignores recent Nature findings that the politically asymmetric sanctions of existing social media content moderation systems are due to differences in misinformation sharing. Crowd-sourced content moderation has a proven track record of working excellently at Wikipedia but might not be so effective at Meta if you can only report the toast for assessment AFTER it is burnt.

Clearly, too, the classification metric of what Zuckerberg terms “illegal and high-severity violations” vis-à-vis “lower-severity violations” has a strong eye-of-the-beholder quality. Nonetheless, Alex Schultz, chief marketing officer at Meta and its highest-ranking gay executive, has already confidently hailed the pending greater volumes of LGBTQ hate speech as potentially progressive.

However, this decision to effectively almost completely move away from content moderation at scale could, perhaps, also be fruitfully understood from the perspective of examining the data about executive cultures underlying over three centuries of business failures.

Our research found that the elite teams delivering business scandals, disasters and strategic missteps adhere to six common dysfunctional archetypes: subordination (lack of independence), imbalance (missing key voices), distension (culture becomes exaggerated or misdirected), bystander (diffusion of responsibility sees key players refuse to speak up or point to everyone else rather than take responsibility), bureaucratic (rule-bound cultures ignore and sideline alternative views, content or ideas), and conformity (groupthink and failure to raise legitimate concerns). You can judge which of these further dysfunctions do or don’t apply to Meta.

Facebook was identified by our research as a prime example of an executive suite suffering from the “lack of independence” board dysfunction. There is structurally no opportunity for departure from whatever Mark Zuckerberg decides goes without community notes, fact checkers and freely expressed dissenting voices. The freer speech that is good for the Meta platform customers moving forward remains forbidden for shareholders and executives alike.

This often happens when the organisation has a dominant chief executive with a strong character but also, in this instance, because of the way power within the organisation is structured. In corporate governance analysis terms, when the chief executive is also a dominant shareholder with priority voting rights, they have always got the power to control appointments to the board. And when this happens, it becomes hard if not impossible for the executive to be subject to proper scrutiny.

Clearly, the subordinated board dysfunction is also simultaneously present at Meta so, despite the clamorous mood music to restore free speech and (re)prioritise differing voices, what we’re looking at here is still a board that’s incapable of acting independently of the executive. If nothing else, it is clear that the Meta board won’t, at least while on duty, be taking part in this “new era” of freer speech.

To be fair, with 40,000 content moderation staff, Meta were until this announcement market-leading investors in and exponents of such editorial supervision practices (though the success of their own algorithms at disseminating misinformation and hate speech at scale meant that it effectively proved beyond them. Now, the abdication of direct corporate responsibility for content moderation at scale that Zuckerberg values can also be viewed through the optic of the ownership and shareholding structures that historically, and still today, apply at Facebook.

These dual-status share class ownership structures have been fashionably de rigueur in the AI and digital start-up and venture capital investment space for most of the twenty-first century, in order to supposedly incentivise ambitious and visionary founders of usually digital and AI-related newly-minted companies. Such privileged shareholdings inevitably enable both swift, inspired decision-making as well as complete unaccountability for founder faux pas, strategic errors, flip-flops, unfettered entrepreneurialism and grandstanding.

This is the third election in a row in which Mark Zuckerberg has issued a cris de coeur and revised fundamentals of the Meta mission in relation to speech and censorship (see John Herrman on the Eternal Apology Tours of 2016, 2021and now 2025).

Mark Zuckerberg makes his strategic decisions in the context of an editorial board that effectively cannot, in practical terms, moderate, gainsay or change any of his decisions – should they wish to or not – precisely because of the way shareholder votes accrue to his preferential share class type compared to those available elsewhere to other shareholders accessing them via public markets. In the case of both old and “new era” Facebook, Zuckerberg wields enormous control over the company’s structure, despite the fact that it’s publicly owned, because of the Facebook dual class structure. Each share of class A stock has one vote, whereas Mark Zuckerberg holds class B stock where each share equals 10 votes.

Historically to, if nothing else, try to manage perceptions pertaining to this shareholding inequality and boardroom inequity, Meta made great play of the standalone supervisory (but not legally binding) fig leaf of its “independent” Oversight Board. Though effectively only an optional extra wheel for the Meta Zuckerberg unicycle, said Oversight Board is supposed to function as a moderating influence to critically examine strategic and/or board-level decisions and, where necessary, highlight any concerns they, markets, users and shareholders may hold or believe germane. Some see this as Meta corporate governance window-dressing but, in light of their immediate full-throated endorsement of these content moderation changes, they can, perhaps, nowadays be better understood as a prime example of the “bystander” psychological archetype commonly found in dysfunctional boards.

With two different boards at Meta limited from effectively holding the chief executive to account, it transpires, if judged by press reports, that this new-era mission of freer speech isn’t permitted for employees. Internal scepticism about the latest Meta replacement board member appointees Dana White, John Elkann and Charlie Songhurst has been throttled by the HR department. These new board directors replace with immediate effect (to use online games jargon) non-playing character Nick Clegg, who joins Sheryl Sandberg as another alumnus to proudly own a Meta severance non-disclosure agreement.

But, more widely, while proactive content moderation is apparently a barrier to free speech in the wider world, it is curious that the external published press reports of internal criticism was quickly moderated from being easily found on Facebook.

So, how can we seek to better understand such elite team decision-making at Meta and else wherein a more well-rounded fashion? Our preferred optic is to examine underlying social and psychological archetypes/dysfunctions, as explained above. Business history is littered with executive groups falling prey to psychological blindspots and errors.

To conclude, while the actual Meta platforms are getting rid of fact-checkers, they already don’t exist at board level. Indeed, Zuckerberg is able to operate Meta and make it strategically flip-flop the way it does precisely because of the various social, structural, shareholding and psychological dysfunctions of its imbalanced board(s). The legitimacy and efficacy of claims about globally permitting fuller freedom of expression on Facebook and Instagram platforms should be seen in the light of its own constrained oversight context as, clearly, the “Let Them Eat Free Speech” banquet is not always on the menu at Meta.


Sign up for our weekly newsletter here


About the author

Gerry Brown

Gerry Brown is Chairman of Novaquest Capital Management.

Randall S Peterson

Randall S Peterson is Professor of Organisational Behaviour and Founding Director of the Leadership Institute at London Business School.

Posted In: Management | Technology

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.