In Silicon Valley, Sam Altman, OpenAI chief executive, once called the "AI crown prince," is facing a wave of fierce criticism. He once championed artificial intelligence (AI) for humanity and drew global expectations, but he is now at the center of harsh attacks as a "fraud" and a "liar," amid competitors' all-out offensives, ruptures with key partners, and consumer backlash. Analysts say OpenAI's solo run is crumbling as not only its technical edge but the moral foundation of the corporations itself is being shaken.
According to the industry on the 5th, the sharpest criticism of OpenAI is coming from former colleagues who once shared the same table. Dario Amodei, CEO of Anthropic, founded by OpenAI's founding members, blasted OpenAI in a recent memo to staff, calling its conduct a "brazen lie" and "gaslighting."
Right after Anthropic rejected the Ministry of National Defense's request for large-scale domestic surveillance and the use of autonomous lethal weapons on ethical grounds, OpenAI promptly signed the contract, prompting the rebuke: "They are only eager to placate employees and have no interest in preventing abuse." Amodei mocked Altman's self-portrayal as a pacifist as "safety theater" and declared a philosophical break with OpenAI.
This morality controversy is translating into declines in concrete business metrics. According to market researcher Sensor Tower, the day after news broke that OpenAI had joined hands with the Ministry of National Defense, the ChatGPT app's deletion rate surged 295% in a single day. Users are engaging in so-called "rating terrorism," leaving one-star ratings for ChatGPT on Apple's App Store, while Anthropic's AI model "Claude," benefiting in turn, climbed to No. 1 among free apps on the U.S. App Store.
As ChatGPT, once "synonymous with AI," begins to be shunned by consumers, some say market leadership is tilting further, with Anthropic's annualized revenue soaring to $19 billion.
Doubts about Altman's "character" and "communication style" appear to be spreading across external relations. In particular, the conflict with Elon Musk, xAI chief executive and the biggest backer at OpenAI's founding, is expected to peak in court on Apr. 27. Musk filed a fraud and breach-of-contract suit, saying "OpenAI broke its nonprofit promise and morphed into a for-profit corporations seeking private gain." In the complaint, Musk said he invested on the premise that OpenAI was a nonprofit research institute but was "manipulated and deceived" as Altman and co-founders engaged in for-profit activities. He also accused "Altman and his accomplices" of betrayal.
Musk argues OpenAI effectively turned into a for-profit corporations by partnering with Microsoft (MS). Microsoft is currently OpenAI's largest investor, reportedly having invested about $13 billion. The clash has already spilled into a public war of words, and in Silicon Valley some say the showdown between two founders representing the AI industry is becoming a symbol of the sector's power dynamics.
Cracks with strategic partners are also concerning. On the 4th (local time), Jensen Huang, Nvidia's CEO, said "additional investment in OpenAI is unlikely," signaling a withdrawal from a previously planned $100 billion investment. SoftBank, chaired by Masayoshi Son, long seen as OpenAI's strongest ally, is also reportedly at odds over "Stargate," a massive data center project, due to a trademark dispute and differences over site selection.
On top of that, Apple's recent choice of Google's "Gemini," not OpenAI, as Siri's partner led OpenAI to declare a "code red (state of emergency)" internally, suggesting the limits of Altman's leadership.
In the industry, the current situation is seen as the result of Altman's pursuit of "speed-at-all-costs" and "opportunistic management." The fact that more than 900 technical staff across Anthropic, Google, and even OpenAI employees have formed a united front against military applications of AI reflects a consensus that Altman's unilateral decisions damaged trust across the industry.
If OpenAI loses the ethical battle over "who can build a more trustworthy AI," beyond technical performance contests, the listing planned for this year and the defense of a $500 billion corporate valuation are likely to face red flags.