52 comments

[ 1.9 ms ] story [ 43.2 ms ] thread
Will lawful use be determined in secret courts a la NSA and FISA?
Who defines "lawful" if Google and the Pentagon disagree?

> The classified deal apparently doesn’t allow Google to veto how the government will use its AI models.

Seems concerning?

It's pretty funny how these guys are all becoming some kind of internet version of, like, Halliburton. It seems pretty desperate. B2C and B2B applications didn't pan out I guess?
When my sister and I would play monopoly as kids, we had lost the manual so whenever we didn’t like the outcome of whatever happened, we would make up rules about what was right. Technically then, it was very easy stay compliant while still being able to do well because we could rewrite the rules.

Also, since I was older I feel like I was able to get away with those redefinitions a lot more often…

The pentagon is part of the executive, not the legislative, and as such they can not write the 'rules' (law)
We need to encourage kids to play and make mistakes more so they can prepare themselves for the real world.

Which is really just a bunch of big children with bank accounts, drugs, and weapons.

How well does this hold up in terms of legal scrutiny when previous actions indicate that the Pentagon would retaliate against Google if they didn't accept this "lawful use only" farce?

Could Google back out of this agreement later by arguing that they were coerced?

Not trying to suggest that Google would be opposed to doing evil, but curious about how solid this agreement would be in practice.

there is 0 reason that the definitions of 'lawful' for the purposes of these agreements should be classified.
And that is news-worthy because unlawful use is normal?
Unsurprising from Google, but still bad. If Google has no right to object to a particular use, this is equivalent in practice to "any use, lawful or not".
Lawful is meaningless in the context of the Trump administration. Should Google waver (which they won't), they'll be declared a supply chain risk or otherwise bullied into submission.
This all works if you assume that any action the government takes must be “lawful”. The assumption here is that the Pentagon is obeying the law and any unlawful use would go through normal reporting / violation channels - same as any illegal order or violation or whistleblower report.

The Pentagon does not want Google or anyone else deciding what they can and cannot use their AI for. They’re saying we won’t break the law, and that should be enough for you - pinky swear!

And that seems to be enough for Google. Though I might request some auditing capability that is agentic to verify rather than take them at their word.

Next step: is Google FEDRAMP’d yet for this and for classified enclaves? Or do they also go through Palantir’s AI vehicle?

> We remain committed to the private and public sector consensus that AI should not be used for domestic mass surveillance or autonomous weaponry without appropriate human oversight.

And starts the lying to our faces. The public and private (from your own employees!) consensus is that it should not be used for those things at all, regardless of “human oversight.”

Huh. I never realized the T-800 runs on Android. Makes sense, I guess.
Do no evil. Well don't make anything illegal at least. I mean, let's not do what is different from whatever we wish at the moment.
Snakes. All of them
One observation.

Having your work being used by the govt in ways you disagree with feels similar to having your taxes used in ways you disagree.

When you pay taxes you have no say in the bombs acquired with that and where they are dropped. The latter though doesn't seem to provoke the same push back