Empire Of Lies? What 250 Openai Insiders Allegedly Told A Journalist About Sam Altman — And Why The World Should Be Worried

Every empire in history has told its subjects the same story: that the conquest is for their own good. Rome called it civilisation. The colonial powers called it enlightenment. Today's artificial intelligence giants call it progress. And if a growing body of insider testimony is to be believed, the newest empire may be the most quietly powerful one yet — built not with soldiers, but with servers.

At the centre of this story sits Karen Hao, an award-winning investigative journalist, former Wall Street Journal correspondent covering American and Chinese technology companies, and co-host of the podcast The Interface. Her reporting career has taken her deep inside the boardrooms of Silicon Valley's most secretive companies, and her latest book, the bestselling "Empire of AI: Inside the Reckless Race for Total Domination," is being widely discussed as one of the most consequential investigations yet published into the inner workings of OpenAI. According to her own account, the book draws on conversations with roughly 250 current and former OpenAI insiders — a scale of access that few outsiders have ever achieved into a company this guarded.

What she reportedly found, if her account holds, deserves the attention of every government, parent, worker, and citizen on earth, because the decisions being made inside a handful of California offices are quietly reshaping the future all of us will live in.

THE AGI MYTH
One of the most provocative claims attributed to Hao's reporting is this: that the term "AGI" — artificial general intelligence — functions less as a scientific milestone and more as a marketing device. A vague, ever-shifting finish line that can be moved whenever convenient, used to justify enormous valuations, trillion-dollar fundraising rounds, and a narrative of inevitability that discourages regulation. If a goal can never be precisely defined, it can never be definitively missed — and that, some critics argue, is precisely the point. Whether one fully agrees with this framing or not, it is a serious argument made by a serious journalist, and it deserves serious debate rather than dismissal.

THE MANUFACTURED ARMS RACE
Equally provocative is the suggestion that the so-called US-China "AI arms race" — the narrative that America must out-build China at any cost or lose civilisation itself — may be, in part, a politically convenient framing rather than a purely technical inevitability. Fear of a rival superpower has, throughout history, proven remarkably effective at loosening regulatory scrutiny and public resistance. If the story convinces enough people that safety concerns are a luxury the West cannot afford, then caution itself starts to look unpatriotic. A nation frightened of losing a race will rarely stop to ask whether the race was worth running in the first place.

THE HUMAN COST BEHIND THE MACHINE
Perhaps the most sobering thread in this reporting concerns the human labour quietly powering these systems from behind the curtain — data annotators and content moderators, often in lower-income countries, reportedly tasked with reviewing some of the most disturbing material imaginable so that chatbots can respond safely to the rest of us. This is not a new criticism of the AI industry broadly, and it echoes concerns raised for years by labour researchers and journalists covering the global content moderation economy. If true at the scale suggested, it raises an uncomfortable question the industry has been slow to answer: whose wellbeing is being quietly spent to purchase the world's convenience?

THE BOARDROOM DRAMA THE PUBLIC NEVER FULLY UNDERSTOOD

The world already knows the bones of this story, even if the full flesh of it remains contested. In November 2023, Sam Altman, OpenAI's chief executive, was abruptly removed by his own board, only to be reinstated within days following intense pressure from staff and investors — a corporate whiplash that stunned the technology world and made headlines globally. The precise internal reasoning behind that ouster has never been fully, publicly settled, and competing accounts continue to circulate about what was really disputed behind those closed doors — questions of safety philosophy, trust, and the direction of the company among them. Hao's reporting is among the accounts that attempts to reconstruct that fog, drawing on insiders who lived through it, though as with any single account of a highly contested event, readers would do well to weigh it against other reporting rather than treat it as final word.

A FAIR HEARING FOR THE OTHER SIDE
Now, honesty demands I steel-man the opposing view, for it is easy to build a villain and harder to build an honest argument.

It is entirely possible that the leaders of these companies genuinely believe, as many have stated publicly, that they are building something meant to benefit humanity broadly, and that the scale of investment reflects genuine belief in the technology's promise rather than pure cynicism. Ambitious founders throughout history have often been simultaneously sincere believers and ruthless operators — the two are not mutually exclusive. It is also worth noting that OpenAI has, according to public reporting, declined to formally participate in or endorse Hao's book, and any single-sourced insider account, however extensively researched, should be read as one serious perspective among several rather than an uncontested verdict.

WHY THIS MATTERS FOR AFRICA AND THE GLOBAL SOUTH

For those of us writing from Accra rather than Silicon Valley, this is not a distant American soap opera. The decisions made inside these companies will determine which jobs survive automation on this continent, whose data trains these systems, and whose labour quietly cleans up the mess behind the curtain. Africa has been a testing ground for foreign ambition before. We would be wise, this time, to watch the boardroom as closely as we watch the technology.

THE QUESTION THAT REMAINS
Every empire eventually meets its historians, and history is rarely kind to power that operated in the dark. Whether OpenAI's story ends in vindication or reckoning, one truth already stands undisputed: the public deserves more transparency than it has been given, and a story this consequential should never depend on the goodwill of the very people it investigates.

The race, we are told, cannot be stopped. Perhaps the more urgent question is whether it was ever meant to be won by the rest of us at all.

Author's note: I write this piece not as an enemy of technology, but as a Ghanaian storyteller and engineering student who believes progress without accountability is simply conquest wearing a friendlier name. History remembers those who asked the uncomfortable questions long before it remembers those who profited from the silence.

About the author: Chief Tutu Baffour Asare Brownsy Williams is a highly acclaimed independent Ghanaian author, columnist, filmmaker, and digital content creator, and the founder of Brownsy Silva Company, a multi-disciplinary creative enterprise spanning literature, film, and digital storytelling. A student of engineering at Accra Technical University, he brings a rare fusion of technical insight and literary depth to his commentary on technology, society, and power. His opinion columns for Modern Ghana are widely read across Ghana, the United Kingdom, the United States, Canada, and Germany, and his creative works span multi-generational novels, screenplays, and short films exploring African identity and the modern world.