OpenAI's inner workings fully exposed: A ten-thousand-word report tears open Sam Altman's true face!
This recently released "OpenAI Files" report, like a depth charge, has completely unearthed all the dark history of OpenAI and Sam Altman.
The report was jointly published by two non-profit tech oversight organizations, Midas Project and Tech Oversight Project, and claims to be the most comprehensive investigation into OpenAI's governance issues to date. Spanning over ten thousand words, it covers company documents, legal lawsuits, open letters, media reports, and various other materials.
Rob Wiblin (@robertwiblin) exclaimed after reading the report: "There's so much crazy stuff in there!"
And the most outrageous part is—
Altman actually falsely claimed to be the chairman of Y Combinator in SEC filings, and he kept this lie going for several years!
The report indicates that when Altman left Y Combinator, he unilaterally proposed to transition from president to chairman, even preemptively publishing this news on the company's website. The problem is, Y Combinator's partners never agreed to this!
The news was later taken down, but Altman continued to use this false title in SEC filings.
This is just the tip of the iceberg.
Elon Musk directly lashed out
After seeing this report, Elon Musk (@elonmusk) unceremoniously tweeted: "Scam Altman."
Musk was blunt, directly calling Sam Altman "Scam Altman."
OpenAI's "profit cap" was actually a joke
Do you remember OpenAI's initial promise of a 100x profit cap?
The original intent of this design was to ensure that if the company truly created AI capable of automating all human labor, the profits would belong to all humanity.
However, the report reveals that this cap has been quietly changed to an annual increase of 20%. At this rate, the cap will exceed 100 trillion dollars in 40 years!
Moreover, OpenAI has never publicly disclosed this modification, continuing to promote its "profit cap" structure.
I have no equity in OpenAI?
Altman adamantly stated in a congressional hearing that he had no equity in OpenAI and only earned enough to cover his health insurance.
But the report reveals that he holds indirect shares through Sequoia Capital and Y Combinator funds.
Interestingly, Altman owns a 7.5% stake in Reddit.
When Reddit announced its partnership with OpenAI, his net worth instantly surged by $50 million. He also invested in Rain AI, and then OpenAI signed a letter of intent to purchase $51 million worth of chips.
Upon seeing this, netizen Smeldor Elcatraz (@Smeldorr) suddenly realized: "This might explain why ChatGPT can return Reddit search results. I thought they had a licensing agreement before."
Former executives collectively exposed: "Altman is not suitable to lead AGI"
Co-founder Ilya Sutskever directly told the board: "I don't think Sam is the one who should control the AGI button."
He even provided a self-destructing PDF file, which contained Slack screenshots documenting "dozens of examples of lying or other toxic behavior."
Former CTO Mira Murati stated: "I am uneasy with Sam leading us towards AGI."
Dario Amodei (Anthropic founder) even used terms like "gaslighting" and "psychological abuse" to describe Altman's management tactics.
At least five other OpenAI executives provided similar negative feedback to the board.
Restrictive agreements and whistleblower threats
The report also reveals OpenAI's strict control over employees.
Departing employees were forced to sign extremely stringent confidentiality and non-disparagement agreements, lifetime prohibitions from criticizing their former employer, and even from acknowledging the existence of the agreement. Refusal to sign or violation of the agreement could result in the forfeiture of all vested equity—
Which could be worth millions of dollars.
OpenAI also required employees to waive their rights to federal whistleblower compensation. Former employees have filed complaints with the SEC, accusing OpenAI of unlawfully preventing them from reporting to regulators.
Terry Hargrove (@terryhargrove) questioned: "Why are there no consequences for lying to Congress now? No consequences for lying to the public?"
Safety culture? Long ago yielded to product
In 2023, OpenAI experienced a major security vulnerability where hackers stole AI technical details, but the company concealed it for over a year before reporting. More ironically, when Leopold Aschenbrenner shared security concerns with the board, OpenAI directly fired him.
Jan Leike, former head of the alignment team, said upon his departure: "Over the past few years, safety culture and processes have given way to shiny products."
Altman's "startup black history"
This is not the first time Altman's leadership has been questioned.
The report shows that at his first startup, Loopt, senior employees twice tried to get the board to fire him, citing "deceptive and chaotic behavior."
However, Rob Wiblin later added that the evidence for this claim is relatively weak; it might be true, but it might also be false.
Restructuring plan: The final step from non-profit to for-profit
The report details OpenAI's ongoing restructuring. The company plans to completely remove investor return limits and transition to a Delaware Public Benefit Corporation (PBC) with common stock.
Although OpenAI claims the non-profit organization still retains control, in reality, the non-profit board may no longer possess all the power truly needed to fulfill its mission. Rumors suggest Altman might acquire a 7% equity stake after the restructuring, valued at approximately $20 billion.
Damian Tatum (@_damian_bot) lamented: "I will only believe these companies can solve AI alignment when they make some progress on human alignment. Such people should not control transformative technology."
Is someone backing this report?
Myke Mnemonic (@findmyke) posted a screenshot asking who funded the report.
However, the report's creators explicitly stated: They have not received funding, assistance, editorial guidance, or any form of support from Musk, xAI, Anthropic, Meta, Google, Microsoft, or any OpenAI competitors.
Altman has not yet responded
XAlphaone (@xalphaone_) directly called out Altman: "Hey @sama, would you care to respond to these accusations? I'd like to hear your perspective. If you ignore it, then it must be true. 🤷♂️ Truth always comes out. Remember God is in control."
As of now, neither Altman nor OpenAI has issued an official response to this report.
The timing of this report's release is particularly delicate—just as OpenAI is preparing to complete a major restructuring and its valuation is soaring to $300 billion. It not only exposes internal chaos within the company but also raises a fundamental question:
What happens when idealism meets $60 billion in venture capital and what might be the most profitable technology in history?
The report concludes by proposing a vision for reform, encompassing responsible governance, ethical leadership, and shared benefits.
Finally, Andrea Elliott (@DreaLeePower) offered some dark humor: "I guess this makes me feel slightly better that they think there will be a society in 40 years from which to extract profits. 😅😰🤦🏼♀️🤷🏼♀️"
What do you think about these allegations concerning Sam Altman?
The Verge report: https://www.theverge.com/openai/688783/the-openai-files-will-help-you-understand-how-sam-altmans-company-works
OpenAI Files full report: https://www.openaifiles.org/
Helen Toner interview evaluation: https://www.lesswrong.com/posts/fJseTLvaun4SmRv4a/an-evaluation-of-helen-toner-s-interview-on-the-ted-ai-show#Claim_6__Senior_managers_at_Loopt_asked_the_board_to_fire_Altman__twice__