Technology

Did we read the same thing?

We support the use of AI for lawful foreign intelligence and counterintelligence missions. But using these systems for mass domestic surveillance is incompatible with democratic values.

So they accept surveillance in other countries? What about other countries’ democratic values?

Even fully autonomous weapons (those that take humans out of the loop entirely and automate selecting and engaging targets) may prove critical for our national defense. But today, frontier AI systems are simply not reliable enough to power fully autonomous weapons.

So you don’t because it still sucks? But if it didn’t, you would?

And what about legal?

• Do Not Develop or Design Weapons???
• Do Not Compromise Privacy or Identity Rights???

I’ve really lost my faith in the US. They think they hold the power, but they’re missing the point: real power is built on trust-and we’re losing more of it every day.

So the government wants "full self-driving" attack drones. You know, just in case the military actually disobeys an unlawful order?

How many pieces of science fiction do we have where the "bad guys" are literally just killer robots we created and then realized we didn't have control over? Why would we decide it is a good idea to literally build terminators? I'm convinced the government will actually build the "orphan crushing machine" next...

And I see the big baby in chief has answered in typical baby fashion.

HN thread broke all this down and pointed out the PR wiggle room.

They'll cave. These companies always do

Anthropic is now playing Good Cop in a charade. They don't care about ethics.

Leaked Slack Messages Show CEO of “Ethical AI” Startup Anthropic Saying It’s Okay to Benefit Dictators

Anthropic’s CEO admits compromising with authoritarian regimes to secure AI funding

Anthropic CEO Dario Amodei backs President Trump on AI policy, pushes back on criticism

"... Without a subscription. For the full, unlocked dictatorship just the low low price of a bajillion dollars a month will give you the power you need to defeat your enemies."

I read somewhere that Anthropic has $18,000,000,000 in commitments from last year alone, so conceivably, they can stand to lose a mere $200,000,000 and it won't create a huge issue for them in the short term.

I hope that's how they're looking at it.

cannot in good conscience

🤣

Amodei "we cannot in good conscience allow this".

Hegseth looks confused, turns towards his team and mouths "...in good what?""

I guess it's good that they draw the line somewhere, but it is absolutely horrifying to me as a non-American that the moral stance is limited to:

• taking issue with fully autonomous AI weapons (purely for technical reasons according to this letter, they are working hard on making them possible)
• refusing to conduct mass surveillance of US citizens specifically (foreign nationals are fair game and the intelligence apparatus will surely only be used for good and to preserve democracy).

This is not Anthropic refusing to cooperate with the Trump administration as the title may suggest, they are in fact explicitly eager to serve the US Department of War. They are just vying for slightly better contract terms.

Department of ~~War~~ Defense

Are those the same AI systems that recommended nuclear escalation in 90% of simulations?

I can't see the name "anthropic" without thinking about furries.

Anthro pic.

Now you can't either. You're welcome.

In other words, they did the calculations and found that they don't yet have the market share or the financial position that would enable them to sell out to the government. However, they're planning to get there someday and hope the DoD is willing to work together in the future.