What's this about?
Altman posted alongside the letter, and the notable part is who he addressed it to: "we are happy if you want to work with us or any of our competitors or partners, but please take this moment seriously." For a CEO whose usual mode is to promote his own product, explicitly telling people to go to a competitor is a deliberate signal about how the labs want this framed.
Why is this framed as "collective"?
Because security has a property most products don't: your defence depends on other people's defence. If the hospital, the water utility and the DNS provider you rely on are compromised, your own hardening does not save you. Economists call this a positive externality — the benefit of one party's investment spills over to everyone. Externalities are systematically under-funded by markets, which is why the letter ends up asking governments to pay.
Why is it worth noticing?
- The framing is a shift: from "our model is the safest" to "the capability is arriving everywhere and the response has to be shared".
- It lands two days after OpenAI's own incident report, which makes the timing read as consequence rather than coincidence.
- Read next to Mollick's warning from yesterday's digest about the pre-open-weights window, the industry is now saying the same thing in two registers.