Sometimes a single old sentence says more about a company than any current press release. Simon Willison dug up an email that came to light in the Musk v. Altman case — written by Sam Altman to OpenAI’s board on October 1, 2022.
The quote
«We have been having extensive discussions around open source strategy. We will discuss it more at our next board meeting, but one thing we’d like to do soon is to create a language model with the approximate capability of GPT-3 that can run locally on consumer hardware and release that. We’d like to do it soon, before Stability or someone else does. In general, we think this helps discourage others from releasing similarly-powerful models, and makes it harder for new efforts to get funded.»
Read that last sentence again. Releasing an open model here wasn’t a commitment to openness. It was a strategic maneuver — a tool to squeeze competitors and make it harder for them to reach capital.
Why this matters right now
The timing is remarkable. This very week, Axios and others report that the Trump administration is considering banning Chinese open-weight models like Kimi K3 at the behest of American frontier labs. And an OpenAI strategist openly called for manufacturing regulatory fear around open models.
So four years after this email, the mindset hasn’t really changed — only the direction. In 2022, an open model was meant to starve the competition. In 2026, a ban on open models is meant to do the same job. The shared pattern: openness is a lever in competition, not an end in itself.
My take
I want to be fair — an internally floated idea from 2022 isn’t a verdict on today’s company, and people and organizations change. But the email is a useful corrective against overly simple narratives. When labs talk about «safety» or «responsibility» in the context of open models, it’s worth asking whose interest is quietly riding along.
For me, it just underlines why the current open-source debate is so charged. It’s rarely only about the technology or the risks. It’s about business models — and about who gets to write the rules in the end. This email is a small window into exactly that kind of thinking. And it’s a good reason to listen closely when the labs reach for the big words.
Sources: