In 2017, Sam Altman told US government officials that China had launched what he called an "AGI Manhattan Project" - a national crash program to build artificial general intelligence, meaning an AI system capable of human-level reasoning across any task, not just the narrow applications that exist today. He used this argument to push for billions in US government funding to keep pace. An intelligence official who reviewed the claim later put it plainly: "It was just being used as a sales pitch."
This account, surfaced in recent investigative reporting, reframes a pivotal early moment in AI funding history. 2017 was early in the current AI cycle - GPT-1 wouldn't be released until 2018, and OpenAI was still operating as a non-profit research lab. But Altman was already working Washington, and the framing he chose - existential competition with China, wartime urgency - would go on to become the dominant justification for government AI investment across the following decade.
What Intelligence Found
Officials who reviewed Altman's claims found no evidence that China had launched any centralized AGI program at that scale. The conclusion wasn't just that he was wrong about China's capabilities at the time. It was that the framing appeared calculated to persuade rather than accurately inform.
A Framing That Never Went Away
The more significant issue isn't a single exaggerated claim made nine years ago. It's that the strategy worked. The existential China competition narrative became a fixture in congressional hearings, executive orders, and AI export control debates. Whether Altman's 2017 pitch contributed directly is hard to isolate, but the rhetorical template - national security urgency, peer adversary with unlimited resources, no time to deliberate - has never left the conversation.
OpenAI has since grown into one of the most valuable private companies in the world, with reported valuations above $300 billion. That trajectory was built on early access to large-scale compute and investment relationships that required someone to make the urgency case in terms decision-makers would respond to.
For people using AI tools today, this is background context, not an operational concern. ChatGPT, DALL-E, and the OpenAI API function independently of their founder's 2017 briefings. But it matters for evaluating AI policy arguments still being made today, many of which rely on the same China-race framing. Knowing that framing was recognized as a sales pitch at the time it was first delivered is relevant context for how much weight to give it now.