Is this the most consequential AI regulation battle we have seen yet?
I don’t care about what you “feel” is right, I do not care about your “moral values”. I want a good clean, policy/material impact (economics, national security, functional democracy etc.) focused debate to know:
- Is it legal / defensible for Hegseth to effectively seize Anthropic if they refuse to let the Pentagon use Claude for 1) increased mass surveillance and 2) fully autonomous armed “operations”?
- What does this mean for the other frontier models? Should we be concerned who decides to say Yes instead (or who might’ve already said “Yes) in case Anthropic decides to stand their ground?
- Does this open up the conversation to nationalize the frontier models (which isn’t uncommon in times of War, for ex: BMW in Germany during WWII)? What would that look like today?
These are just some starter questions I’m wrestling with myself, so feel free to respond freely to any / all / none of them. You WILL lose brownie points if I can sniff you used AI to respond. This is not a safe space for your offshored brain….More context:
——
From Jenni Griffin, Chief National Security Correspondent for FOX, on X:
At high stakes Pentagon meeting today Sec Hegseth gave Anthropic head Dario Amodei ultimatum to allow the Pentagon to use Anthropic’s AI model for mass domestic surveillance and kinetic autonomous operations without human oversight or face censure and be labeled “supply chain threat.”
According to a source familiar:
The meeting was cordial, not a dressing down, not a screaming match, all business.
Hegseth praised the Anthropic product but then said if by Friday Anthropic does not agree to the Pentagon’s use of the model without restrictions, then Hegseth would terminate the contract and use the Defense Production Act to force Anthropic to comply AND/OR designate Anthropic a supply chain threat and national security risk. (EDIT: Both are mutually exclusive. You can’t be a supply chain risk but also invoke the DPA to say that the country needs this product so much for national security that it will override any restrictions put in place by the company that limits govt access to the product. Both cannot be true.)
At issue is Anthropic’s two stipulations that its advanced AI model currently used in the Pentagon’s classified systems is NOT used for autonomous kinetic operations (Anthropic currently requires human oversight of autonomous operations when used to kill things for safety reasons because they don’t know how the autonomous system will react and could even endanger soldiers using the model; soldiers and others could lose control of the model and automatically start killing large groups without humans in the “kill chain.”) Second Anthropic bars its models from being used for mass domestic surveillance. Hegseth wants these restrictions lifted.
According to a source familiar with the talks, Anthropic has never objected to the use of its models for “legitimate military operations.” It also told the Hegseth it never complained to the Pentagon or Palantir about the use of its models in the Maduro raid.
——
come ahhn…nobody? (Monkey bot repellent)
If you want a debate on serious policy, that almost definitionally excludes anything the Trump Administration has done or plans to do. So you've answered your own question. Case closed.
I'm sorry I don't have more to contribute to this conversation, but I do get uncomfortable when the government tried to force private companies to do anything. I'm not saying it should never be done (e.g., I do think tik tok needed to be figured out as it was basically a large scale Chinese data aggregation exercise).
In this particular case I think Anthropic should win, but someone may eventually say yes and that it won't really matter.
You're overstating -- the seizing is a tail risk scenario (more realistically they just get termianted), and seizing an old version of the model is going to be pointless as frontier model innovation continues
If Anthropic disagrees, they can just go their own away. It's entirely their right. Just as it's the Pentagon's right to have a maximally reliable vendor because the consequences of not having this are incredibly dire
No one on the side of "mass domestic surveillance" is ever the good guy. It really is that simple and it didn't use to be a "left v. right" issue.
Besides, Hegseth needs Anthropic to help fake his bench press videos.
Agreed on it not being (or shouldn't be) a left or right issue. It does always seems that the people in power seem to like it (see Bush starting the patriot act, Obama/Trump/Biden/Trump all continuing/expanding it), and the people out of power start talking about government overreach while most of the citizenry are against it.
The Trump administration listed Anthropic as a supply chain risk (typically reserved for companies in China, Russia, etc. that we can't have anywhere in our military for obvious reasons) because they didn't bow to their demands. Anthropic is not a company with close ties to China or Russia and there's no reason a Lockheed engineer or government worker should be banned from using Claude to help with their day to day. If anything, Anthropic could make our government more efficient and more competitive with China and such.
Just another example of Trump blatantly bullying private companies to make them do as the government says and the so called "free market, small government crowd" saying "yes daddy".
Do not want the government using AI to mass surveil or kill people as I imagine (hopefully) 80%+ Americans agree with.
I don't understand the overly punitive punishment on supply chain risk etc. If they don't want to do it, it's their right as an American company and just cut the contract and find someone else. Classic Trump admin BS from his fucktard loyalist administration.
Voluptas deserunt dolor doloribus est culpa ad. Sit fugiat sed error harum reiciendis. Sunt est voluptatem sint enim. Aut quidem est odio pariatur.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...