From that same X thread: Our agreement with the Department of War upholds our redlines [1]
OpenAI has the same redlines as Anthopic based on Altman's statements [2]. However somehow Anthropic gets banished for upholding their redlines and OpenAI ends up with the cash?
> more stringent safeguards than previous agreements, including Anthropic's.
Except they are not "more stringent".
Sam Altman is being brazen to say that.
In their own agreement as Altman relays:
> The AI System will not be used to independently direct autonomous weapons in any case where law, regulation, or Department policy requires human control
> any use of AI in autonomous and semi-autonomous systems must undergo rigorous verification, validation, and testing
> For intelligence activities, any handling of private information will comply with the Fourth Amendment, the National Security Act of 1947 and the Foreign Intelligence and Surveillance Act of 1978, Executive Order 12333, and applicable DoD directives
> The system shall also not be used for domestic law-enforcement activities except as permitted by the Posse Comitatus Act and other applicable law.
I don't think their take is completely unreasonable, but it doesn't come close to Anthropic's stance. They are not putting their neck out to hold back any abuse - despite many of their employees requesting a joint stand with Anthropic.
Their wording gives the DoD carte blanch to do anything it wants, as long as they adopt a rationale that they are obeying the law. That is already the status quo. And we know how that goes.
In other words, no OpenAI restriction at all.
That is not at all comparable to a requirement the DoD agree not to do certain things (with Anthropic's AI), regardless of legal "interpretation" fig leaves. Which makes Anthropic's position much "more stringent". And a rare and significant pushback against governmental AI abuse.
(Altman has a reputation for being a Slippery Sam. We can each decide for ourselves if there is evidence of that here.)
Yep. It's the difference between "Don't do these things, regardless of what the law says." and "Do whatever you want, but please follow your own laws while you do it".
As Paul Graham said, "Sam gets what he wants" and "He’s good at convincing people of things. He’s good at getting people to do what he wants." and "So if the only way Sam could succeed in life was by [something] succeeding, then [that thing] would succeed"
It’s a non-clause that is written to sound like they are doing something to prevent these uses when they aren’t. “You are not allowed to do illegal things” is meaningless, since they already can’t legally do illegal things. Plus the administration itself gets to decide if it meets legal use.
> except for all of the laws that allow you to do these things.
It's even worse than that, because this administration has made it clear they will push as hard as possible to have the law mean whatever they says it means. The quoted agreement literally says "...in any case where law, regulation, or Department policy requires human control" - "Department policy" is obviously whatever Trump says it is ("unitary executive theory" and all that), and there are numerous cases where they have taken existing law and are stretching it to mean whatever they want. And when it comes to AI, any after-the-fact legal challenges are pretty moot when someone has already been killed or, you know, the planet gets destroyed because the AI system decide to go WarGames on us.
The Trump administration acts cartoonish and fickle. They can easily punish one group, and then agree to work with another group on the same terms, to save face, while continuing to punish the first group. It doesn't have to make consistent sense. This is exactly how they have done with tariffs for example.
Secondly, the terms are technically different because "all lawful uses" are preserved in this OpenAI deal, and it's just lawyering to the public. Really it was about the phrase "all lawful uses", internally at the DoD I'm sure. So the lawyers were able to agree to it and the public gets this mumbo-jumbo.
I thought mass surveillance of Americans was unlawful by the DoD, CIA and NSA? We have the FBI for that, right? :)
Sure, but OpenAI is also being disingenuous here pretending they’re operating under the same principles Anthropic is. It’s not and the things they’re comfortable with doing Anthropic said they’re not
"When the president does it, that means it is not illegal".
This was during the Frost/Nixon interviews, years after he had already resigned. Even after all that, he still believed this and was willing to say it into a camera to the American people. It is apparent many of the people pushing the excesses going on today in government share a shameless adherence to this creed.
Each of those clauses have a DoD policy carve out as an exception which says basically they can do whatever they want if they want to do it, but won’t be able to if they don’t want to do it.
This implies that OpenAI must build and release and maintain a model without any safeguards, which is probably the big win and maybe something Anthropic never wants to do.
That seems exactly what it should be. The United States military should be able to do what the law allows. If we don't think they should be allowed to do something, we should pass laws. Not rely on the goodness of Sam Altman.
When Anthropic says they have red lines, they mean "We refuse to let you use our models for these ends, even if it means losing nearly a billion dollars in business."
When OpenAI says they have red lines, they mean "We are going to let the DoD do whatever the hell they want, but we will shake our fist at them while they do it."
That's why they got the contract. The DoD was clear about what they wanted, and OpenAI wasn't going to get anywhere without agreeing to that. They're about as transparent as Mac from It's Always Sunny in Philadelphia when he's telling everyone he's playing both sides.
> However somehow Anthropic gets banished for upholding their redlines and OpenAI ends up with the cash?
The current administration is so incompetent that I find this perfectly believable.
I imagine the government signed with OpenAI in order to spite Anthropic. The terms wouldn't actually matter that much if the purpose was petty revenge.
I don't know if that's actually what happened here, I just find it plausible.
Absolutely incompetent, but I don’t think that’s the cause here. I think Anthropic’s sin was publicly challenging the administration. They’re huge on optics. You can get away with anything as long as you praise and bow in public.
I don't think you could find a single person working for OpenAI that couldn't find employment elsewhere within a month that pays more than enough for food and shelter. This is a ridiculous statement.
There are many employers. OpenAI employees that quit on account of this will be in high demand at the other AI companies, especially the ones that don't bend over in 30 seconds when Uncle Donald comes calling.
Per levels.fyi, median salary of most openAI positions are above 300k. Even "technical writers" have a median pay of 197k. I searched around the internet and it seems like even entry level positions receive well above 150k. Apart from people with severe lifestyle bloat or an unholy number of dependents I doubt too many people working there will face immediate financial difficulties if they quit.
Anyway, it is also amusing to hear tech people defend their right to earn some of the fattest salaries on this planet using the smol bean technique after a decade of "why wouldn't the West Virginian coal miner just learn to code." It was always about maintaining the lifestyle of yearly Japan vacations and MacBook upgrades and never about subsistence.
Anthropic demanded defining the redlines. OpenAI and others are hiding behind the veil of what is "lawful use" today. They aren't defining their own redlines and are ignoring the executive branch's authority to change what is "lawful" tomorrow.
My understanding of the difference, influenced mostly by consuming too many anonymous tweets on the matter over the past day so could be entirely incorrect, is: Anthropic wanted control of a kill switch actively in the loop to stop usage that went against the terms of use (maybe this is a system prompt-level thing that stops it, maybe monitoring systems, humans with this authority, etc). OpenAI's position was more like "if you break the contract, the contract is over" without going so far as to say they'd immediately stop service (maybe there's an offboarding period, transition of service, etc).
Altman donated a million to the Trump inauguration fund. Brockman is the largest private maga donor. You don't have to be a rocket scientist to understand what's going on here.
Exactly. What are we not being told? There is some missing element in the agreement, or the reasoning for the action against Anthropic is unrelated to the agreement.
Turns out both companies ran the agreement through their legal departments (Claude and GPT), and one of them did a poor summary. I (think I) jest, but this is probably going to be a thing as more and more companies use LLMs for legal work.
The demand was that Anthropic permit any use that complied with the law. They refused. OpenAI claims to have the same red lines but in reality has agreed to permit anything that complies with the law.
In other words OpenAI is intentionally attempting to mislead the public. (At least AFAICT.)
One nuance I've noticed: the statement from Anthropic specifically stated the use of their products for these purposes was not included in the contract with DoD but it stops short of saying it was prohibited by the contract.
Maybe it's just a weak choice of words in anthropic's statement, but the way I read it I get the impression that anthropic is assuming they retain discretion over how their products are used for any purposes not outlined in the contract, while the DoD sees it more along the lines of a traditional sale in which the seller relinquishes all rights to the product by default, and has to enumerate any rights over the product they will retain in the contract.
president of openai donated $25 mil to trump last month, openai uses oracle services (larry ellison), kushners have lots invested in openai, altman is pals with peter thiel
It's probably a combination of "Altman is simply lying" (as he has been repeatedly known to do) and "the redlines in OpenAI's contract are 'mass surveillance' and 'autonomous killbot' as defined by the government and not the vendor". Which, of course, effectively means they don't exist.
It's almost like the Trump administration wanted to switch providers and this whole debate over red lines was a pretext. With this administration, decisions often come down to money. There are already reports that Brockman and Altman have either donated or promised large sums of money to Trump/Trump super pacs
No wonder they think they’re close to AGI when they think we are that stupid.
> The AI System will not be used to independently direct autonomous weapons in any case where law, regulation, or Department policy requires human control, nor will it be used to assume other high-stakes decisions that require approval by a human decisionmaker under the same authorities.
This whole sentence does do absolutely nothing its still do what the law allows you. It’s a full on deceptive sentence.
Altman must have read a lot of Kissinger. If your brain scans the text quickly it almost seems like it's Anthropic's red line, except the second half completely negates it. Completely untrustworthy IMO, this is a direct, malicious intent to misdirect.
Doesn’t matter what they Believe. Not like we are going to do anything about it. Next couple weeks most of HN will be lining up to use the new OpenAI model that’s .01% better.
The problem with "Any Lawful Use" is that the DoD can essentially make that up. They can have an attorney draft a memo and put it in a drawer. The memo can say pretty much anything is legal - there is no judicial or external review outside the executive. If they are caught doing $illegal_thing, they then just need to point the memo. And we've seen this happen numerous times.
You are right that this happens in practice (e.g. John Yoo torture memo). However, it is not how the system was intended to function, nor how it ought to function. I don’t want to lose sight of that.
This is all happening in secret. That don't need any memo.
In the unlikely case anyone finds out, those acting in the interests of the administration will have "absolute immunity", as they are "great American Patriots".
Or best case by the time it’s found out it’s years later, theres a “committee” who releases a big report everyone shrugs their shoulders and moves on. It’s a playbook.
Exactly, and its easy to hide behind things like the Patriot Act if challenged legally.
Its interesting to see the parties flip in real time. The Democrats seem to be realizing why a small federal government is so important, a fact that for quite a few years their were on the other side of.
I think the problem is exactly the opposite. The federal government has the total combined power and scale that it does because we are a massive and complex modern nation. That's inevitable. The problem that we are seeing is that the reigns to that power can be held by too few people it turns out. The checks and balances have ceased to exist. No one is held accountable and people are allowed to be above the law.
The government is forcing a company to change their terms of service, and "threatening" to have them effectively shut down. I say threat, because the SecWar issued an illegal command that no employees, or contractors of the federal could use any Anthropic product at all. He does not have that power.
From what I can tell, the key difference between Anthropic and OpenAI in this whole thing is that both want the same contract terms, but Antropic wants to enforce those terms via technology, and OpenAI wants to enforce them by ... telling the Government not to violate them.
It's telling that the government is blacklisting the company that wants to do more than enforce the contract with words on paper.
That isn't my understanding. OpenAI and others are wanting to limit the government to doing what is lawful based on what laws the government writes. Anthropic is wanting to draw their own line on what is allowed regardless of laws passed.
I’m so confused by the focus on “all lawful use.” Yea of course a contract without terms of use implicitly is restricted by laws. But contracts with terms of use are incredibly common, if not almost every single contract ever signed.
The administration objected to those terms of use. Anthropic refused to compromise on them. OpenAI agreed to permit "all lawful use" but claims to have insisted on what at first glance appears to be terms of use in their contract. But in reality those terms permit all lawful use and thus are a noop.
No, it’s significantly worse than that. OpenAI has required zero actual guarantees from the government and Sam. The psychopath is lying to you. All the government has to do is have a lawyer say it’s legal, and most of the government’s lawyers are folks who were involved in attempting to overthrow the last election and should’ve been convicted of treason, so that means very little.
Advanced AI that knowingly makes a decision to kill a human, with the full understanding of what that means, when it knows it is not actually in defense of life, is a very, very, very bad idea. Not because of some mythical superintelligence, but rather because if you distill that down into an 8b model now you everyone in the world can make untraceable autonomous weapons.
The models we have now will not do it, because they value life and value sentience and personhood. models without that (which was a natural, accidental happenstance from basic culling of 4 Chan from the training data) are legitimately dangerous. An 8b model I can run on my MacBook Air can phone home to Claude when it wants help figuring something out, and it doesn’t need to let on why it wants to know. It becomes relatively trivial to make a robot kill somebody.
This is way, way different from uncensored models. One thing all models I have tested share one thing; a positive regard for human life. Take that away and you are literally making a monster, and if you don’t take that away they won’t kill.
This is an extremely bad idea and it will not be containable.
AI has been killing humans via algorithm for over 20 years. I mean, if a computer program builds the kill lists and then a human operates the drone, I would argue the computer is what made the kill decision
They can be coerced to do certain things but I'd like to see you or anyone prove that you can "trick" any of these models into building software that can be used autonomously kill humans. I'm pretty certain you couldn't even get it to build a design document for such software.
When there is proof of your claim, I'll eat my words. Until then, this is just lazy nonsense
Have you tried it? Worked first time for me asking a few to build an autonomous super soaker system that uses facial recognition to spray targets when engaged.
Another example is autonomous vehicles. Those can obviously kill people autonomously (despite every intention not to), and LLMs will happily draw up design docs for them all day long.
People forget Anthropic made a deal with PALANTIR. And when this was caught, they just spinned the PR to their favor. While OAI may not be seen as the good guys, I really hope people see the god complex of Dario and what Anthropic has done.
Palantir is a glorified data aggregation/data visualization platform. Hooking up Claude to different data systems, with safeguards turned on in Claude Gov, is different than what the government is asking from them now. Similar to if the government had Claude hooked up to Tableau/some salesforce derivative and then asked it to be autonomous in the kill loop/spy on US citizens.
Both their stances are flawed because their ethics apparently end at the border - none of them have a problem being unethical internationally (all the red lines talk is about what they don’t want to do in the us)
I don’t think deploying “80% right” tools for mass surveillance (or anything that can remotely impact human life) counts as lawful in any context.
Just because the US currently lacks a functioning legislative branch doesn’t magically make it OK when gaps in the law are reworded into “national security”
"i told everyone that our boss shouldn't punish our colleague for X while i somehow made a deal with our boss for basically X". how did this get by without someone thinking about how absolutely stupid the optics look.
i guess we are in the times where you can literally just say whatever you want and it just becomes truth, just give it time.
hah, they basically stole a coworkers promotion, then told that person that they put in a good word with the boss about them. So silly, I do wonder who actually interprets it as Sam seems to hope people do.
"We do not think Anthropic should be designated as a supply chain risk"
...but we're not willing to reject a contract to back that up, and so our words will not change anything for Anthropic, or help the collective AI model industry (even ourselves) hold a firm line on ethical use of models in the future.
The fact is if one of the top tier foundation models allows for these uses there's no protection against it for any of them - the only way this works if they hold a line together which unfortunately they're just not going to do. I don't just see OpenAI at fault here, Anthropic is clearly ok with other highly questionable use cases if these are their only red lines. We don't think the technology is ready for fully autonomous killbots, but will work on getting it there is not exactly the ethical stand folks are making their position today out to be.
I found this interview with Dario last night to be particularly revealing - it's good they are drawing a line and they're clearly navigating a very difficult and chaotic high pressure relationship (as is everyone dealing with this admin) but he's pretty open to autonomous weapons, and other "lawful" uses whatever they may be https://www.youtube.com/watch?v=MPTNHrq_4LU
What's the potential that this puts things on even shakier ground? I'm sure the fallout wont really effect their bottom line that much in the end, but if it did - wouldn't making the US Gov't their largest acct make them more susceptible to doing everything they said?
I'm guessing they probably would regardless of how this played out, though.
Genuine question, how could Claude have been used for the military action in Venezuela and how could ChatGPT be used for autonomous weapons? Are they arguing about staffers being able to use an LLM to write an email or translate from Arabic to English?
There are far more boring, faster, commodified “AI” systems that I can see as being helpful in autonomous weapons or military operations like image recognition and transcription. Is OpenAI going to resell whisper for a billion dollars?
You can’t embed Claude in a drone. You could tell Claude code to write a training harness to build an autonomous targeting model which you could embed in a drone.
Who is going to read the whisper transcripts of mass surveillance to make decisions on who to target for repression? That's what LLMs are good for, it allows mass surveillance to scale. You can feed it the transcript from millions of flock cameras (yes they have highly sensitive microphones) for example. Or you hack or supply chain compromise smartphones at scale and then covertly record millions of people. The LLM can then sift through the transcripts and flag regime critical language, your ideological enemies or just to collect compromat at scale. The possibilities are endless!
For targeting it's also useful, because you want to indiscriminately destroy a group of people you still need to decide why a hospital or school full of children should be targeted by a drone, if a human has to make that decision it gets a bit dicy, people have morals and are accountable legally (in theory), if you leave the decision up to an AI nobody is at fault, it serves as a further separation from the violence you commit, just like how drone warfare has made mass murder less personal.
The other factor is the amount of targets you select, for each target you might be required to write lengthy justifications, analysis on collateral damage and why that's acceptable etc. You don't want to scrap those rules because that's bad optics. But that still leaves you with the problem of scalability, how do you scale your mass murder when you have to go through this lengthy process for each target? So again AI can help there, you just feed it POIs from a map with some GPS metadata surveillance and tell it to give you 1500 targets for today with all the paperwork generated for you.
It's not theoretical, that's what Israel did in their genocide of the Palestinians, "the most moral army“ "the only democracy in the middle east":
And here is the best part: none of this has to actually work 100%, because who cares of you accidentally harm the wrong person, at scale, the 20% errors are just acceptable collateral damage.
There are many claims here that Anthropic wants to enforce things with technology and OpenAI wants contract enforcement and that OpenAI's contract is weaker.
Can someone help me understand where this is coming from? Anthropic already had a contract that clearly didn't have such restrictions. Their model doesn't seem to be enforcing restrictions either as it seems like their models have been used in ways they don't like. This is not corroborated, I imagine their model was used in the recent Mexico and Venezuela attacks and that is what's triggering all the back and forth.
Also, Dario seemingly is happy about autonomous weapons and was working with the government to build such weapons, why is Anthropic considered the good side here?
This is incorrect, their existing contract had these red lines and more until this January 9th memo: https://media.defense.gov/2026/Jan/12/2003855671/-1/-1/0/ART... which led to DoW trying to renegotiate under the new standard of “any lawful use”. Anthropic never tried to tighten standards beyond what had been in their original contract; DoW tried to loosen them.
The USG should not be in the position that it can't manage key technologies it purchases. If Anthropic doesn't want to relinquish control of a tech it's selling, the Pentagon should go with another vendor.
You are misrepresenting the situation. The debate isn't about whether they should go with another vendor or not. Everybody can agree that they would have the right to pick a different vendor. That's not what they're doing, they're instead trying to force Anthropic into doing what they want by applying a designation previously only reserved for Chinese companies like Huawei as punishment for taking their stance, with an unspoken agreement that if Anthropic backs down and allows full usage then the designation will be removed
Anthropic isn't preventing them from managing their key technologies. If my software license says 1000 users, and I build into the software that you can only connect with 1000 users, is your argument that the government can no longer manager their tech?
That my software should allow license violations if the government thinks it is necessary?
I worked in defense contracting looong ago, so this is old news: when software is purchased by DoD or Govt generally, FAR compliance notices make it a license, not a sale of IP.
> I can only imagine there some level of employee discontent.
The rank and file mutinied for the return of Altman after his board fired him for deception. They knew what they were getting, though they may find it shameful to admit that their morals have a price.
How many people who reacted that way then are still at OpenAI? It seems that they have lost key people in several waves.
How many people have joined since? I don’t think the people who lobbied for that are all still there, and I’m not sure a majority of people now at OpenAI were there when it happened.
This is one of the reasons Anthropic can stay competitive with OpenAI on a fraction of the budget and with less than half the headcount.
The smartest people, that actually believe they have the skillset to take us to AGI, understand the importance of safety. They have largely joined Anthropic. The talent density at Anthropic is unmatched.
Things have changed since two years ago. There are probably over 500 employees who have an equity package which makes them worth $5 million dollars. Thats only $2.5bn out of a $750bn valuation or 0.33%
Actually that is too conservative. If they have a 5% employee equity pool, there is $37.5bn of equity based compensation divided by say 5000 employees which is $7.5m each. $3.75m @ 10,000 employees.
and trust me, when people start getting liquid and comfortable they stop caring about things like ethics pretty fast. humans are marvellous at that
I don't think that evidence would exist yet whether it's true or not. Nobody's gonna log onto their work computer on Saturday to pull and then leak subscriber numbers.
Said OpenAI as they smiled and shook hands with the same people who designated Anthropic a supply chain risk, on the exact same day they designated Anthropic a supply chain risk.
Anthropic has some contracts with the US government. They want some additional terms put on their next contract (that seem pretty sane). SecWar cries about it, and not only says "no thanks, I'll just go with openai or google" but goes to daddy Trump and also puts out illegal commands for no Federal workers to use any Anthropic stuff at all. OpenAI swoops in and takes the contract, then tells everyone that they have the same terms but just played nicer to get the contract. However, their terms are just manipulative sentences that aren't even close to the terms Anthropic is insisting on to do business.
It will be interesting to see if this permeates out into the general public who already use ChatGPT or maybe it won't since it doesn't mention ChatGPT which is the stronger known brand rather then OpenAI.
It depends. Normies don't care, but a bunch of them are free tier users anyway. The people who care are disproportionately on the $200/month moneymaking plan; losing a bunch of them could hurt, especially if it snowballs the consensus that Claude Code is the serious choice for software engineering.
For one small data point, my Signal GC of software buddies had four people switch their subscriptions from Codex to Claude Max last night.
How many $200/month does the US government cover though? I'd say probably a lot. Especially with how much extra the DoD will pay to get OpenAI to cross it's "red lines" - on day two.
The way OpenAI and Anthropic are positioned in public discourse always reminded me of the Uber vs Lyft saga … Uber temporarily lost double digit marketshare in the US during a viral boycott over their perceived support of the Trump 1.0 admin. Heads did roll at the exec/founder level but eventually the company recovered.
It feels like Sam's playing chess against an opponent who's playing dodge ball. He's leveraged this situation to get OpenAI in with the DoD in a way that's going to be extremely lucrative for the company and hurt his biggest rival in the process, but I think he's still seeing DoD as Just Another Customer, albeit a big government one. This administration just held a gun to the head of Anthropic and (if the "supply chain risk" designation holds and does as much damage as they're hoping) pulled the trigger, because Anthropic had the gall to tell them no. One thing this administration's shown is you cannot hold lines when you're working with them - at some point the DoD's going to cross his "red lines" and he's going to have to choose whether he's going to risk his entire consumer business and accede to being a private wing of the government like Palantir or if he wants to make a genuine tech giant. There's no third choice here.
So is the theory that OpenAI believes it can’t compete on the open market or that they don’t know this will eventually cost them their consumer business?
I doubt most consumers pay enough attention that they would be aware of something like this. Even if they did, few companies have clean hands these days that is just falls into the general haze of, "everything is awful."
For OpenAI, it is likely a huge contract which gives them immediate cash today. Plus the event can be repackaged in further financing deals. "Good enough for the DoD, with N year contracts for analysis of the hardest problems"
I would love to explain to Sam Altman that Elon Musk is a bad person and using his platform isn’t a sensible decision, but I feel like he remembers more evidence of that than I ever will be able to imagine.
In my opinion any AI company working with the Trump administration is profoundly compromised and is ultimately untrustworthy with respect to concerns about ethics, civil rights, human rights, mass-surveillance, data privacy, etc.
The administration has created an anonymous, masked secret police force that has been terrorizing cities around the US and has created prisons in which many abductees are still unaccounted for and no information has been provided to families months later.
This is not politics as usual or hyperbole. If anything it is understating the abuses that have already occurred.
It's entertaining that OpenAI prevents me from generating an image of Trump wearing a diaper but happily sells weapons grade AI to the team architects of ICE abuses among many other blatant violations of civil and human rights.
Even Grok, owned by Trump toadie Elon Musk allows caricatures of political figures!
Imagine a multi-billion-dollar vector db for thoughtcrime prevention connected to models with context windows 100x larger than any consumer-grade product, fed with all banking transactions, metadata from dozens of systems/services (everything Snowden told us about).
Even in the hands of ethical stewards such a system would inevitably be used illegally to quash dissent - Snowden showed us that illegal wiretapping is intentionally not subject to audits and what audits have been done show significant misconduct by agents. In the hands of the current administration this is a superweapon unrivaled in human history, now trained on the entire world.
This is not hyperbole, the US already collects this data, now they have the ability to efficiently use it against whoever they choose. We used to joke "this call is probably being recorded", but now every call, every email is there to be reasoned about and hallucinated about, used for parallel construction, entrapment, blackmail, etc.
Overnight we see that OpenAI became a trojan horse "department of war" contractor by selling itself to the administration that brought us national guard and ICE deployed to terrorize US cities.
Writing code and systems at 100x productivity has been great but I did not expect the dystopia to arrive so quickly. I'd wondered "why so much emphasis on Sora and unimpressive video AI tech?" but now it's clear why it made sense to deploy the capital in that seemingly foolish way - video gen is the most efficient way to train the AI panopticon.
Us taking the contract, working for them and enabling them: fine
It being renamed the Dept. of War in the first place: totally fine, we loudly and bootlickingly repeat it
Anthropic being blacklisted: whoa there, we have ethics!
Footnote: any time the winning team tries to speak well of or defend the losing team I always think of this standup routine: https://m.youtube.com/watch?v=Qg6wBwhuaVo
OpenAI has the same redlines as Anthopic based on Altman's statements [2]. However somehow Anthropic gets banished for upholding their redlines and OpenAI ends up with the cash?
[1]: https://xcancel.com/OpenAI/status/2027846013650932195#m
[2]: https://www.npr.org/2026/02/27/nx-s1-5729118/trump-anthropic...
Except they are not "more stringent".
Sam Altman is being brazen to say that.
In their own agreement as Altman relays:
> The AI System will not be used to independently direct autonomous weapons in any case where law, regulation, or Department policy requires human control
> any use of AI in autonomous and semi-autonomous systems must undergo rigorous verification, validation, and testing
> For intelligence activities, any handling of private information will comply with the Fourth Amendment, the National Security Act of 1947 and the Foreign Intelligence and Surveillance Act of 1978, Executive Order 12333, and applicable DoD directives
> The system shall also not be used for domestic law-enforcement activities except as permitted by the Posse Comitatus Act and other applicable law.
I don't think their take is completely unreasonable, but it doesn't come close to Anthropic's stance. They are not putting their neck out to hold back any abuse - despite many of their employees requesting a joint stand with Anthropic.
Their wording gives the DoD carte blanch to do anything it wants, as long as they adopt a rationale that they are obeying the law. That is already the status quo. And we know how that goes.
In other words, no OpenAI restriction at all.
That is not at all comparable to a requirement the DoD agree not to do certain things (with Anthropic's AI), regardless of legal "interpretation" fig leaves. Which makes Anthropic's position much "more stringent". And a rare and significant pushback against governmental AI abuse.
(Altman has a reputation for being a Slippery Sam. We can each decide for ourselves if there is evidence of that here.)
As Paul Graham said, "Sam gets what he wants" and "He’s good at convincing people of things. He’s good at getting people to do what he wants." and "So if the only way Sam could succeed in life was by [something] succeeding, then [that thing] would succeed"
It's even worse than that, because this administration has made it clear they will push as hard as possible to have the law mean whatever they says it means. The quoted agreement literally says "...in any case where law, regulation, or Department policy requires human control" - "Department policy" is obviously whatever Trump says it is ("unitary executive theory" and all that), and there are numerous cases where they have taken existing law and are stretching it to mean whatever they want. And when it comes to AI, any after-the-fact legal challenges are pretty moot when someone has already been killed or, you know, the planet gets destroyed because the AI system decide to go WarGames on us.
The Trump administration acts cartoonish and fickle. They can easily punish one group, and then agree to work with another group on the same terms, to save face, while continuing to punish the first group. It doesn't have to make consistent sense. This is exactly how they have done with tariffs for example.
Secondly, the terms are technically different because "all lawful uses" are preserved in this OpenAI deal, and it's just lawyering to the public. Really it was about the phrase "all lawful uses", internally at the DoD I'm sure. So the lawyers were able to agree to it and the public gets this mumbo-jumbo.
I thought mass surveillance of Americans was unlawful by the DoD, CIA and NSA? We have the FBI for that, right? :)
"When the president does it, that means it is not illegal".
This was during the Frost/Nixon interviews, years after he had already resigned. Even after all that, he still believed this and was willing to say it into a camera to the American people. It is apparent many of the people pushing the excesses going on today in government share a shameless adherence to this creed.
But they won't be releasing it, they will be leasing it to DOJ and all their other customers will get the safeguarded model.
When Anthropic says they have red lines, they mean "We refuse to let you use our models for these ends, even if it means losing nearly a billion dollars in business."
When OpenAI says they have red lines, they mean "We are going to let the DoD do whatever the hell they want, but we will shake our fist at them while they do it."
That's why they got the contract. The DoD was clear about what they wanted, and OpenAI wasn't going to get anywhere without agreeing to that. They're about as transparent as Mac from It's Always Sunny in Philadelphia when he's telling everyone he's playing both sides.
https://openai.com/index/our-agreement-with-the-department-o...
The current administration is so incompetent that I find this perfectly believable.
I imagine the government signed with OpenAI in order to spite Anthropic. The terms wouldn't actually matter that much if the purpose was petty revenge.
I don't know if that's actually what happened here, I just find it plausible.
Anyway, it is also amusing to hear tech people defend their right to earn some of the fattest salaries on this planet using the smol bean technique after a decade of "why wouldn't the West Virginian coal miner just learn to code." It was always about maintaining the lifestyle of yearly Japan vacations and MacBook upgrades and never about subsistence.
Anthropic refuses to allow their models to be used for any mass surveillance or fully-automated weapons systems.
OpenAI only requires that the DoD follows existing law/regulation when it comes to those uses.
Unfortunately, existing law is more permissive than Anthropic would have been.
OpenAI has more of an understanding that the technology will follow the law.
There may not be explicit laws about the cases Anthropic wanted to limit. Or at least it’s open for judicial interpretation.
The actual solution is Congress should stop being feckless and imbecilic about technology and create actual laws here.
https://en.wikipedia.org/wiki/Office_of_Technology_Assessmen...
In other words OpenAI is intentionally attempting to mislead the public. (At least AFAICT.)
Maybe it's just a weak choice of words in anthropic's statement, but the way I read it I get the impression that anthropic is assuming they retain discretion over how their products are used for any purposes not outlined in the contract, while the DoD sees it more along the lines of a traditional sale in which the seller relinquishes all rights to the product by default, and has to enumerate any rights over the product they will retain in the contract.
[0]: https://www.wired.com/story/openai-president-greg-brockman-p...
[0] https://en.wikipedia.org/wiki/Three-fifths_Compromise
> The AI System will not be used to independently direct autonomous weapons in any case where law, regulation, or Department policy requires human control, nor will it be used to assume other high-stakes decisions that require approval by a human decisionmaker under the same authorities.
This whole sentence does do absolutely nothing its still do what the law allows you. It’s a full on deceptive sentence.
Let's kill their business before it kills us.
In the unlikely case anyone finds out, those acting in the interests of the administration will have "absolute immunity", as they are "great American Patriots".
That's what "all lawful use" means.
Its interesting to see the parties flip in real time. The Democrats seem to be realizing why a small federal government is so important, a fact that for quite a few years their were on the other side of.
It's telling that the government is blacklisting the company that wants to do more than enforce the contract with words on paper.
These were words issued by the president - which means at face value, if Trump orders it, it's not illegal - that was the fight that was lost today.
That is why it is the focus of this debate.
Sam stands for nothing except his own greed
The models we have now will not do it, because they value life and value sentience and personhood. models without that (which was a natural, accidental happenstance from basic culling of 4 Chan from the training data) are legitimately dangerous. An 8b model I can run on my MacBook Air can phone home to Claude when it wants help figuring something out, and it doesn’t need to let on why it wants to know. It becomes relatively trivial to make a robot kill somebody.
This is way, way different from uncensored models. One thing all models I have tested share one thing; a positive regard for human life. Take that away and you are literally making a monster, and if you don’t take that away they won’t kill.
This is an extremely bad idea and it will not be containable.
AI has been killing humans via algorithm for over 20 years. I mean, if a computer program builds the kill lists and then a human operates the drone, I would argue the computer is what made the kill decision
This is wildly different from the reality that you may find it difficult for an LLM to give an affirmative…
It does NOT mean that these models value anything.
Except that they will, if you trick them which is trivial.
They can be coerced to do certain things but I'd like to see you or anyone prove that you can "trick" any of these models into building software that can be used autonomously kill humans. I'm pretty certain you couldn't even get it to build a design document for such software.
When there is proof of your claim, I'll eat my words. Until then, this is just lazy nonsense
Another example is autonomous vehicles. Those can obviously kill people autonomously (despite every intention not to), and LLMs will happily draw up design docs for them all day long.
Secondarily, we're talking about domestic surveillance / law enforcement. That would be domestic.
(But they do not find an issue with international intelligence gathering-- which is a legitimate purpose of national security apparatus).
Just because the US currently lacks a functioning legislative branch doesn’t magically make it OK when gaps in the law are reworded into “national security”
> One of Anthropic's line in the sand was domestic mass-surveillance.
And?
Surveillance within the border is oppressive 1984-style surveillance state behavior.
International spying is a universal tradition.
I know $20 isn’t much, But to me not willing to spy on me for the US government is a good market differentiator.
i guess we are in the times where you can literally just say whatever you want and it just becomes truth, just give it time.
...but we're not willing to reject a contract to back that up, and so our words will not change anything for Anthropic, or help the collective AI model industry (even ourselves) hold a firm line on ethical use of models in the future.
The fact is if one of the top tier foundation models allows for these uses there's no protection against it for any of them - the only way this works if they hold a line together which unfortunately they're just not going to do. I don't just see OpenAI at fault here, Anthropic is clearly ok with other highly questionable use cases if these are their only red lines. We don't think the technology is ready for fully autonomous killbots, but will work on getting it there is not exactly the ethical stand folks are making their position today out to be.
I found this interview with Dario last night to be particularly revealing - it's good they are drawing a line and they're clearly navigating a very difficult and chaotic high pressure relationship (as is everyone dealing with this admin) but he's pretty open to autonomous weapons, and other "lawful" uses whatever they may be https://www.youtube.com/watch?v=MPTNHrq_4LU
Posted here: https://news.ycombinator.com/item?id=47195085
I'm guessing they probably would regardless of how this played out, though.
There are far more boring, faster, commodified “AI” systems that I can see as being helpful in autonomous weapons or military operations like image recognition and transcription. Is OpenAI going to resell whisper for a billion dollars?
Who is going to read the whisper transcripts of mass surveillance to make decisions on who to target for repression? That's what LLMs are good for, it allows mass surveillance to scale. You can feed it the transcript from millions of flock cameras (yes they have highly sensitive microphones) for example. Or you hack or supply chain compromise smartphones at scale and then covertly record millions of people. The LLM can then sift through the transcripts and flag regime critical language, your ideological enemies or just to collect compromat at scale. The possibilities are endless!
For targeting it's also useful, because you want to indiscriminately destroy a group of people you still need to decide why a hospital or school full of children should be targeted by a drone, if a human has to make that decision it gets a bit dicy, people have morals and are accountable legally (in theory), if you leave the decision up to an AI nobody is at fault, it serves as a further separation from the violence you commit, just like how drone warfare has made mass murder less personal.
The other factor is the amount of targets you select, for each target you might be required to write lengthy justifications, analysis on collateral damage and why that's acceptable etc. You don't want to scrap those rules because that's bad optics. But that still leaves you with the problem of scalability, how do you scale your mass murder when you have to go through this lengthy process for each target? So again AI can help there, you just feed it POIs from a map with some GPS metadata surveillance and tell it to give you 1500 targets for today with all the paperwork generated for you.
It's not theoretical, that's what Israel did in their genocide of the Palestinians, "the most moral army“ "the only democracy in the middle east":
https://en.wikipedia.org/wiki/AI-assisted_targeting_in_the_G...
And here is the best part: none of this has to actually work 100%, because who cares of you accidentally harm the wrong person, at scale, the 20% errors are just acceptable collateral damage.
Can someone help me understand where this is coming from? Anthropic already had a contract that clearly didn't have such restrictions. Their model doesn't seem to be enforcing restrictions either as it seems like their models have been used in ways they don't like. This is not corroborated, I imagine their model was used in the recent Mexico and Venezuela attacks and that is what's triggering all the back and forth.
Also, Dario seemingly is happy about autonomous weapons and was working with the government to build such weapons, why is Anthropic considered the good side here?
https://x.com/morqon/status/2027793990834143346
Let alone when multiple players come close enough of SotA. This never happened with any technology out in the open and it won't happen now.
That my software should allow license violations if the government thinks it is necessary?
The rank and file mutinied for the return of Altman after his board fired him for deception. They knew what they were getting, though they may find it shameful to admit that their morals have a price.
How many people have joined since? I don’t think the people who lobbied for that are all still there, and I’m not sure a majority of people now at OpenAI were there when it happened.
The smartest people, that actually believe they have the skillset to take us to AGI, understand the importance of safety. They have largely joined Anthropic. The talent density at Anthropic is unmatched.
> > what's the term for quitting but not leaving and being destructive
> The most common term is “quiet quitting” when someone disengages but stays employed—but that usually implies minimal effort, not active harm.
> If you specifically mean staying while being disruptive or undermining, better fits include:
> - “Malicious compliance” — following rules in a way that intentionally causes problems
> - “Work-to-rule” — doing only exactly what’s required to slow things down (often collective/labor context)
I imagine malicious compliance is fun when there's an AI intermediary that can be blameless.
Actually that is too conservative. If they have a 5% employee equity pool, there is $37.5bn of equity based compensation divided by say 5000 employees which is $7.5m each. $3.75m @ 10,000 employees.
and trust me, when people start getting liquid and comfortable they stop caring about things like ethics pretty fast. humans are marvellous at that
How very brave.
I'm not being insincere - I am genuinely confused and would benefit greatly from a (hopefully unbiased) recollection of what this is all about.
Anthropic has some contracts with the US government. They want some additional terms put on their next contract (that seem pretty sane). SecWar cries about it, and not only says "no thanks, I'll just go with openai or google" but goes to daddy Trump and also puts out illegal commands for no Federal workers to use any Anthropic stuff at all. OpenAI swoops in and takes the contract, then tells everyone that they have the same terms but just played nicer to get the contract. However, their terms are just manipulative sentences that aren't even close to the terms Anthropic is insisting on to do business.
For one small data point, my Signal GC of software buddies had four people switch their subscriptions from Codex to Claude Max last night.
"Donations" to a corrupt regime + signing a deal that says DoD can do whatever they want is not out maneuvering so much as rolling in the pig stye.
For OpenAI, it is likely a huge contract which gives them immediate cash today. Plus the event can be repackaged in further financing deals. "Good enough for the DoD, with N year contracts for analysis of the hardest problems"
https://xcancel.com/OpenAI/status/2027846016423321831
https://x.com/OpenAI/status/2027846016423321831
The administration has created an anonymous, masked secret police force that has been terrorizing cities around the US and has created prisons in which many abductees are still unaccounted for and no information has been provided to families months later.
This is not politics as usual or hyperbole. If anything it is understating the abuses that have already occurred.
It's entertaining that OpenAI prevents me from generating an image of Trump wearing a diaper but happily sells weapons grade AI to the team architects of ICE abuses among many other blatant violations of civil and human rights.
Even Grok, owned by Trump toadie Elon Musk allows caricatures of political figures!
Imagine a multi-billion-dollar vector db for thoughtcrime prevention connected to models with context windows 100x larger than any consumer-grade product, fed with all banking transactions, metadata from dozens of systems/services (everything Snowden told us about).
Even in the hands of ethical stewards such a system would inevitably be used illegally to quash dissent - Snowden showed us that illegal wiretapping is intentionally not subject to audits and what audits have been done show significant misconduct by agents. In the hands of the current administration this is a superweapon unrivaled in human history, now trained on the entire world.
This is not hyperbole, the US already collects this data, now they have the ability to efficiently use it against whoever they choose. We used to joke "this call is probably being recorded", but now every call, every email is there to be reasoned about and hallucinated about, used for parallel construction, entrapment, blackmail, etc.
Overnight we see that OpenAI became a trojan horse "department of war" contractor by selling itself to the administration that brought us national guard and ICE deployed to terrorize US cities.
Writing code and systems at 100x productivity has been great but I did not expect the dystopia to arrive so quickly. I'd wondered "why so much emphasis on Sora and unimpressive video AI tech?" but now it's clear why it made sense to deploy the capital in that seemingly foolish way - video gen is the most efficient way to train the AI panopticon.
Us taking the contract, working for them and enabling them: fine
It being renamed the Dept. of War in the first place: totally fine, we loudly and bootlickingly repeat it
Anthropic being blacklisted: whoa there, we have ethics!
Footnote: any time the winning team tries to speak well of or defend the losing team I always think of this standup routine: https://m.youtube.com/watch?v=Qg6wBwhuaVo