[piker]

Sorry, I think we may be talking past each other. The language you quoted is an objective standard. If, for example, a court ruled that the government had violated the Constitution using the tool, that language would be breached. I don’t think anything I’ve seen (though we haven’t seen the whole agreement!) allows the government to use the product in violation of the law. Anthropic wanted to go further by further limiting the uses in specific cases.

[_alternator_]

Ok I think we are largely in agreement, though perhaps missing the main point: Anthropic wanted restrictions above and beyond “all legal uses”. This was widely reported in the last few days.

OpenAI is passing off their deal as providing additional safeguards beyond “all legal uses” but the language they’ve released doesn’t seem to support that narrative. I’m incensed, and am attempting to point out the hypocrisy in the hopes that OAI gets some blowback for this cynical stunt.

[piker]

Ok, but I thought my analysis was pretty clear on that point:

> OpenAI acceded to demands that the US Government can do whatever it wants that is legal.