I don't see how OpenAI employees who have signed the We Will Not Be Divided letter can continue their employment there in light of this. Surely if OpenAI had insisted upon the same things that Anthropic had, the government would not have signed this agreement. The only plausible explanation is that there is an understanding that OpenAI will not, in practice, enforce the red lines.
It's comforting to know that some of the brightest minds of our generation are going to work at OpenAI, then quitting a few months later horrified, only to post a short mysterious tweet warning everyone of the dangers ahead. So much for alignment and serving humanity.
> Surely if OpenAI had insisted upon the same things that Anthropic had, the government would not have signed this agreement.
But they did.
"Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement."
The difference is that Anthropic wanted to reserve the right to judge when the red lines are crossed, while OpenAI will defer to the DoD and its policies for that. In both cases, the two parties can claim to agree on the principles, but when push comes to shove, who decides on whether the principles are violated differs.
Seems Anthropic did not understand the questions they were asked. From the WaPo:
>A defense official said the Pentagon’s technology chief whittled the debate down to a life-and-death nuclear scenario at a meeting last month: If an intercontinental ballistic missile was launched at the United States, could the military use Anthropic’s Claude AI system to help shoot it down?
>It’s the kind of situation where technological might and speed could be critical to detection and counterstrike, with the time to make a decision measured in minutes and seconds. Anthropic chief executive Dario Amodei’s answer rankled the Pentagon, according to the official, who characterized the CEO’s reply as: You could call us and we’d work it out.
>An Anthropic spokesperson denied Amodei gave that response, calling the account “patently false,” and saying the company has agreed to allow Claude to be used for missile defense. But officials have cited this and another incident involving Claude’s use in the capture of Venezuelan leader Nicolás Maduro as flashpoints in a spiraling standoff between the company and the Pentagon in recent days. The meeting was previously reported by Semafor.
I have a hunch that Anthropic interpreted this question to be on the dimension of authority, when the Pentagon was very likely asking about capability, and they then followed up to clarify that for missile defense they would, I guess, allow an exception. I get the (at times overwhelming) skepticism that people have about these tools and this administration but this is not a reasonable position to hold, even if Anthropic held it accidentally because they initially misunderstood what they were being asked.
"It’s the kind of situation where technological might and speed could be critical to detection and counterstrike"
Missile detection and decision to make a (nuclear) counterstrike are 2 different things to me but apparently the department of war wants both, so it seems not "just" about missile detection.
> The difference is that Anthropic wanted to reserve the right to judge when the red lines are crossed, while OpenAI will defer to the DoD and its policies for that.
I'm sure it's a matter of interpretation. Anthropic thinks the DoW's demands will lead to mass surveillance and auto-kill bots. The DoW probably disagrees with that interpretation, and all OpenAI needs to do is agree with the DoW.
My bet is that what the DoW wants is pretty clearly tied to mass surveillance and kill-bots. Altman is a snake.
Why do you choose to call it the "DoW"? Its official name is the Department of Defense, it was titled that way by Congress and only Congress can change it. What is your motivation in using a term that the current administration has started to use? Do you also use the Gulf of America when referrring to the body of water that defines the southern edge of the USA?
If your starting position is already that Sam Altman lies about everything that doesn't fit your preconceived positions, that doesn't seem like a very useful meaningful position to update.
Yes, what is implied in this episode is that all big companies that do AI development or provide computing for Ai are now signing for these very shady uses of their technologies.
>Surely if OpenAI had insisted upon the same things that Anthropic had, the government would not have signed this agreement.
Have we been watching the same Trump admin for the last year? That sound exactly like something the government would do: pointlessly throw a fit and end up signing a worse deal after blowing up all political capital.
While that thought crossed my mind, someone in a sub thread of parent comment made a point: OpenAI made a statement about how "We insisted this be not be used in those ways and DoD totally says they won't". Which sounds to me like they ceded any hard terms oand conditions and are letting the DoD use it in "any lawful means" which is what Anthropic didn't stand for.
Nah. It's possible that the agreement still supports the required terms.
There is more to this story behind the scenes. The government wanted to show power and control over our companies and industries. They didn’t need those terms for any specific utility, they wanted to fight “woke” business that stood up to them.
Supposedly OpenAI had the same terms as Anthropic (according to SamA). Maybe they offered it cheaper and that’s why they agreed. Maybe it’s all the lobbying money from OpenAI that let the government look the other way. Maybe it’s all the PR announcements SamA and Trump do together.
"we put them into our agreement." is strange framing is Altman's tweet. Makes me think the agreement does mention the principles, but doesn't state them as binding rules DoD must follow.
I prescribe literally zero truth value to what Sam says. He will say whatever he needs to get ahead. It is honestly irritating to me that you and many others here seem to implicitly assume his messages are correlated with truth, doing his social engineering work for him, as if his word should adjust your priors even slightly.
I don't necessarily think he's lying, but there's so much obvious incentive for him to lie here (if only because his employees can save face).
He doesn't even need to be lying, the comment is vague and contains enough loopholes that it could be true yet meaningless. I explained some that I noticed here: https://news.ycombinator.com/item?id=47190163
And fired from YC for lying. And lied to investors about how many Loopt employees he had. And lied about having 100x the actual number of users when he sold it.
It's this simple: Trump is a criminal. Larry Ellison is his pal. Sam Altman has a huge deal for cloud services from Oracle. Trump is using the DoD budget to backstop Ellison's business.
This is pretty much on the right take on it, although it's much more than that. It's very clear at this point, especially the first conclusion, but people insist in looking to the other side.
Attempting to kneecap the breakout front runner of the major American AI companies to ensure the shittier, politically compliant one wins in the short term? Gee I wonder.
For better or worse, outright nationalization of military related companies is common on a global scale. I plan to do my best to ensure this is a domestic catastrophe, and I hope we'll succeed, but I don't expect other countries to care much about varying levels of regime alignment between two billionaire American defense contractors.
Maybe Sam Altman said nicer things about Donald Trump. Maybe he promised that he would not revoke their API keys when Hegseth directs the military to seize ballots. Maybe he's jockeying for position to take over the government when AGI hits.
Ultimately, I don't know how much the specific reasons matter. Pete Hegseth must be removed from office, OpenAI must be destroyed for their betrayal of the US public, that's all there is to it.
If you're unhappy with this, an immediate way to signal it is with your wallet. In my case I've just uninstalled chatgpt from my phone, cancelled my subscription and will up my spend with anthropic.
The little respect I had left for Sam is now wiped. Makes me sick.
Growing up I always thought AI would be this beautiful tool, this thing that opens the gates to a new society where work becomes optional in a way. But I failed to think about human greed.
I remember following OpenAI way back when it was a non profit explaining how AI uncontrolled could be highly detrimental. Now Sam has not only taken that non profit and made it for-profit. It seems he’s making the most evil decisions he can for a buck.
Cancel your subscription, tell your friends to. And vote to heavily tax these companies and their leaders.
Perfect timing - Had already cancelled my Claude sub over their OAuth ban in external tools and was about to pick up a Codex sub as the next best alternative.
Ended up renewing my Claude sub today instead. Principled stances matter and I no longer trust OpenAI to be trustworthy custodians of my AI History.
Just canceled my subscription! I immediately received an email with the subject “We’d love your feedback on why you canceled your ChatGPT plus subscription” and a link to a survey.
Totally agree. Signed up for a claude code account and will not give OpenAI any money in the future. Let's see what Google does. I will definitely vote with my wallet.
Personally I'm happy about this. OpenAI are being fair about letting the gov use their models to spy on everybody, doesn't seem right that Americans get a pass.
Do you honestly believe that cancelling a subscription makes a bit of difference to a company that is either committing accounting fraud on a monumental scale or shoveling venture capital money into a furnace? not to mention the whole collaborating with a fascist government thing.
taking real action is your choice, but stop pretending this kind of thing matters one iota
edit: to be clear, i'm not advocating for nihilism, but tricking yourself into thinking you made a difference to make yourself feel better isn't the play either
It absolutely matters, especially when done in unison like this.
Cancelling ChatGPT sends a signal that you don't agree with weaponizing AI. Switching to Claude says you support Anthropic's principled stance against it. If you have a strong opinion either way, today is the day to vote with your wallet.
Dismissing every small action as meaningless is just apathy and how nothing ever changes.
Anthropic isn't against weaponizing AI, it's just against two specific carve outs for now. They happily accepted the Pentagon's money so long as it was only spying on other countries. And now that the leopard is eating their face they're claiming the moral high ground.
It's entirely possible for both Anthropic and OpenAI to be in the wrong here. This is a massive publicity win but it doesn't make them heroes in my book.
What has an impact is cancelling a subscription and then talking about it. The media will amplify it the pushback. The goal is to make the name OpenAI and ChatGPT toxic, that whatever you do will be converted into a technology that will surveil or bomb you.
Do they? What are those OpenAI earnings that you are talking about? That's a company that should have ceased existing some time ago if earnings were important
> but stop pretending this kind of thing matters one iota
This is blatantly false and intellectually dishonest. Of course it matters. Your edit is also wrong; you are advocating for nihilism with statments like these.
It's the only thing that matters. These companies don't follow the rules of capitalism physics. They live or die on vibes alone and the tech community abandoning them en masse is bad for the vibes. Once they lose the vibes they are Wiley Coyote looking down at the canyon below.
I think you have too much pessimism. It's not guaranteed to work, but as I mentioned in another thread, since around December, Claude (and Gemini to a lesser extent) has had all the buzz in tech circles, while Chat-GPT has seemed like the also-ran. And that matters: decision-makers in companies notice these things and momentum becomes self-reinforcing (you use Claude Code because everyone else uses Claude Code). If a large enough group of developers visibly defects from OpenAI because of this, it definitely could have consequences. It's not a sure thing, but it's far from hopeless.
I was not a Chat-GPT user even before this, but I'm bumping my Claude Code subscription to the next tier up. Fuck OpenAI.
My knee-jerk reaction to this was looks like an opportunistic maneuver that Sam is known for and I'm considering canceling my subscriptions and business with OpenAI
But what's the most charitable / objective interpretation of this?
Does it suggest that determination of "lawful use" and Dario's concerns falls upon the government, not the AI provider?
Other folks have claimed that Anthropic planned to burn the contentious redlines into Claude's constitution.
Update: I have cancelled my subscriptions until OpenAI clarifies the situation. From an alignment perspective Anthropic's stand seems like the correct long-term approach. And at least some AI researchers appear to agree.
As people have repeatedly mentioned, if the War Department was unhappy with Anthropic's terms, they could have refused to sign the contract. But they didn't: they were fine with it for over a year. And if they changed their mind, they could've ended the contract and both sides could've walked away. Anthropic said that would've been fine. But that's not what happened either: they threatened Anthropic with both SCR designation and a DPA takeover if Anthropic didn't agree to unilateral renegotiation of terms that the War Department had already agreed were fine.
It's absurd, and doubly so if OAI's deal includes the same or even similar redlines to what Anthropic had.
That means autonomous killbots are a-ok. Human responsibility is not the same as human decision-making.
The president or anybody at DoD can be "responsible", and we know there will be zero accountability. The courts defer to the executive, and Congress is all-too-happy for the executive to take the flak for their wars.
>human responsibility for the use of force, including for autonomous weapon systems
So there’s the difference, and an erasure of a red line. OpenAI is good with autonomous weapon systems. Requiring human responsibility isn’t saying much. Theres already military courts, rules of engagement, and international rules of war.
More details on the difference between the OpenAI and Anthropic contracts from one of the Under Secretaries of State:
>The axios article doesn’t have much detail and this is DoW’s decision, not mine. But if the contract defines the guardrails with reference to legal constraints (e.g. mass surveillance in contravention of specific authorities) rather than based on the purely subjective conditions included in Anthropic’s TOS, then yes. This, btw, was a compromise offered to—and rejected by—Anthropic.
> For the avoidance of doubt, the OpenAI - @DeptofWar contract flows from the touchstone of “all lawful use” that DoW has rightfully insisted upon & xAI agreed to. But as Sam explained, it references certain existing legal authorities and includes certain mutually agreed upon safety mechanisms. This, again, is a compromise that Anthropic was offered, and rejected.
> Even if the substantive issues are the same there is a huge difference between (1) memorializing specific safety concerns by reference to particular legal and policy authorities, which are products of our constitutional and political system, and (2) insisting upon a set of prudential constraints subject to the interpretation of a private company and CEO. As we have been saying, the question is fundamental—who decides these weighty questions? Approach (1), accepted by OAI, references laws and thus appropriately vests those questions in our democratic system. Approach (2) unacceptably vests those questions in a single unaccountable CEO who would usurp sovereign control of our most sensitive systems.
> It is a great day for both America’s national security and AI leadership that two of our leading labs, OAI and xAI have reached the patriotic and correct answer here
It is a great day for both America’s national security and AI leadership that two of our leading labs, OAI and xAI have reached the patriotic and correct answer here
He's an administration official openly cheerleading his team. This should be characterized as the insider perspective/spin, not a neutral analysis of the relevant facts.
A government promise that they'll only do lawful things is not reassuring at all:
1. We've seen government lawyers write memos explaining why such-and-such obviously illegal act is legal (see: torture memo). Until challenged, this is basically law.
2. We've seen government change the law to make whatever they want legal (see: patriot act)
3. We've seen courts just interpret laws to make things legal
A contractor doesn't realistically have the power to push back against any of these avenues if they agree to allow anything legal.
(At the risk of triggering Godwin's Law, remember that for the most part the Holocaust was entirely legal - the Nazi's established the necessary authorization. Just to illustrate that when it comes to certain government crimes, the law alone is an insufficient shield.)
You're quoting social media posts from a regime official who says he didn't participate in these negotiations and doesn't work for the relevant department.
If his characterization of the agreement is correct, which I will not believe and you should not believe until a trustworthy news outlet publishes the text, I suppose this would convince me that Hegseth does not literally plan to build a Terminator for democracy-ending purposes. There's a lot of inexcusable stuff here regardless, but perhaps merely boycotting OpenAI and the US military would be a sufficient response if this all checks out.
You're expecting logic from the Trump administration and that's not really how they do things. Maybe it was never about the redlines? Maybe they decided Anthropic was their enemy, and that was their excuse.
Just uninstalled the app and canceled subscription. OpenAI can't justify their insane valuation without an user base. Especially when there are capable models elsewhere.
I bet Sam secretly pledged to DoD that the red lines were only temporary, for optics and to calm employees at the all hands meeting.
A few months down the line, OpenAI will quietly decide that their next model is safe enough for autonomous weapons, and remove their safeguard layer. The mass surveillance enablement might be an indirect deal through Palantir.
That's very possible! In the last few days Anthropic was getting a lot of attention, and OpenAI was looking weaker in comparison. It seems like a politically coordinated job to remove competition.
For sure, he's been pissed that OpenAI no longer has the Mandate of Heaven and Claude is all anyone has been talking about since December. (And it's not just an ego thing: because OAI isn't profitable yet, they need the hype to keep going to raise money on favorable terms, so loss of buzz is an existential threat). I absolutely believe that he started making calls to try and get buddies in the White House to take Anthropic down.
I don't trust Sam to be telling the truth. It would be to his benefit to lie about this and make Anthropic look bad, so he of course would, even if it's not actually the case.
It makes sense if you imagine the real motivation is “make sure the AI contracts go to my good friend Sam”, and all the red line stuff is just a way to pick a fight with Anthropic.
I had kept my Plus subscription just because I was lazy, and it was inexpensive and convenient… but this turn definitely helped me get off the fence. I am exporting and deleting my data now, and the cancellation is already done.
"Starting today I will be asking prominent members of the tech community to sign their name onto this. A code of conduct, authored by me, that pledges them to a universal ethos, which I created, that I call tech ethics or Tethics for short."
Also, it's probably tricky to find a Schelling point that a broad range of people can agree to.
* no military use
* no lethal use
* no use in support of law enforcement
* no use in support of immigration enforcement
* no use in mass surveillance
* no use in domestic mass surveillance (but mass surveillance of foreigners is OK)
* no use in domestic surveillance
* no use in surveillance
* require independent audits
* require court oversight
* require company to monitor use
* require company to monitor use and divulge it to employees
* some other form of human rights monitoring or auditing
* some other form of restriction on theaters/conflicts/targets
* company will permit some of these uses (not purport to forbid them by license, contract, or ToS) but not customize software to facilitate them
* company can unilaterally block inappropriate uses
* company can publicly disclose uses it thinks are inappropriate
* some other form of remedy
* government literally has to explain why some uses are necessary or appropriate to reassure people developing capabilities, and they have some kind of ongoing bargaining power to push back
It feels normal to me that a lot of people would want some of those things, but kind of unlikely that they would readily agree on exactly which ones.
I even think there's a different intuition about the baseline because one version is "nobody works on weapons except for people who specifically make a decision to work for an arms company because they have decided that's OK according to their moral views" (working on weapons is an abnormal, deliberate decision) and another version is "every company might sell every technology as part of a weapons system or military application, and a few people then object because they've decided that's not OK according to their moral views" (refusing to work on weapons is an abnormal, deliberate decision). I imagine a fair number of people in computing fields effectively thought that the norm or default for their industry was the latter, because of the perception that there are "special" military contractors where people get security clearances and navigate military procurement processes, and most companies are not like that, so you were not working on any form of weapon unless you intentionally chose to do so. But, having just been to the Computer History Museum earlier this week, I also see that a lot of Silicon Valley companies have actually been making weapons systems for as long as there has been a Silicon Valley.
There is definitely a muddle on so many levels about signaling and agreeing on ethics in technology.
But as innovation slows globally, it is implementation, ethics, and ideology that will once again be the dominant metrics of progress, so there's a new window emerging to push for this social/moral change in technology once again.
So it's still critically important that we actively work towards finding a meaningful, socially contagious differentiator other than "ethical technologist" even if it's difficult- look at what OpenAI gets away with under that flimsy banner.
I would put bets on the issue probably being that it was pointed out that Anthropic's models were used to assist the raid in Venezuela, Anthropic then aggressively doubled down on their rules/principles and the DOD didn't like being called out on that so they lashed out, hard.
If theres anything this admin doesn't like, its being postured against or called out by literally anyone, especially in public.
I'm unsure how to feel about this whole dust-up. It doesn't seem like much has changed in substance. Maybe OpenAI outmaneuvered Anthropic behind the scenes. Possibly Anthropic was seen as not behaving deferentially enough towards the government. But this administration has proven comically corrupt, so it wouldn't surprise me if money was involved. Will be interested to see what journalists turn up.
It’s amazing how quickly the players keep shifting here.
Yesterday and the day before sentiment seemed to be focused on “Anthropic selling out”, then that shifted to “Anthropic holds true to its principles in a David vs Goliath” and “the industry will rally around one another for the greater good.” But suddenly we’re seeing a new narrative of “Evil OpenAI swoops in to make a deal with the devil.”
Reminds of that weekend where Sam Altman lost control of OpenAI.
He’s certainly solidified his place in the history of this era.
But I suspect the public sentiment will eventually turn against him. When society sets its pitchforks on big tech he’ll be the poster boy. A 21st century John D. Rockefeller.
I considered that comparison, but in all seriousness, I’m not sure it’s apt.
Are he and his peers Hitler or they the naive oligarchs who think they can keep populist leaders and their constituencies under their thumb? Only to be out maneuvered by the people who the masses think have their back.
I know many folks who think their political leaders have the best interest at heart (rightly or wrongly). I know nobody who thinks tech leaders do. At best they want to be them.
So they agreed to the exact same clauses that Anthropic put forward but with OpenAI instead?
So it wasn't about those principles making them a supply chain risk? They're just trying to punish Anthropic for being the first ones to stand firm on those principles?
Anthropic would probably not renegotiate in a year about the principles, while Sam Altman is known to be morally flexible so OpenAI will almost surely allow the military to do what they want in the future. Sam Altman might even have said behind closed doors that these restrictions will be removed once the drama has died down.
For the people that don't understand how they got a deal with the same redlines, it probably because OpenAI agreed to not question them. The safeguards are there, both parties agree now fuck off and let us use your model how we see fit.
Anthropic probably made the mistake of questioning the Military's activities related to Claude after the Venezuela mission and wanted reassurance that the model wouldn't be used for the redlines, and the military didn't like this and told them we aren't using your models unless you agree to not question us and then the back and forth started.
In the end, we will probably have both OpenAI and Anthropic providing AI to the military and that's a good thing. I don't think they will keep the supply chain risk on Anthropic for more than a week.
This is really about the imminent strike on Iran which is now super telegraphed. They are gonna use ChatGPT for target selection, and the likely outcome is that it will fuck things up and a bunch of civilians are going to die because of this decision.
When this happens, Altman will go from being merely a drifter to having blood on his hands.
So does this mean that OpenAI will give whatever the DoD asks for and they will pinky swear that it won’t be used for mass surveillance and autonomous killing machines?
and we know we can trust openAI because they were founded on "open" and "safe" AI (up until they realized how much money there was to be made, at which point their only value changed to "make money")
> The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
(1) Well, did both sides sign the agreement and is it actually effective? Or is it still sitting on someone's desk until it can get stalled long enough?
(2) What does "agreement" even mean? Is it a legally enforceable contract, or just some sort of MoU or pinkie promise?
(3) If it's a legally enforceable contract, is it equally enforceable on all of their contracts, or just some? Do they not have existing contracts this would need to apply to?
(4) What does "reflects them in law and policy" even mean? Since when does DoW make laws, and in what sense do their laws reflect whatever the agreement was? Are these laws he can point to so everyone else can see? Can he at least copy-paste the exact sentences the government agreed to?
1. There's no substantive change. Hegseth/Trump just wanted to punish Anthropic for standing up to them, even if it didn't get them anything else today -- establishing a chilling effect for the future has some value for them in this case, after all. And OpenAI was willing to help them do that, despite earlier claiming that they stood behind Anthropic's decisions.
2. There is a substantive change. Despite Altman's words, they have a tacit understanding that OpenAI won't really enforce those terms, or that they'll allow them to be modified some time in the future when attention has moved on elsewhere.
Either way, it makes Altman look slimy, and OpenAI has aligned with Trump against Anthropic in a place where Anthropic made a correct principled stand. It's been clear for a while that Anthropic has more ethics than OpenAI, but this is more naked than any previous example.
If you believe that any country should have a military and intelligence apparatus, the job of that apparatus is to kill people and surveil foreigners. I do think the US government should have a military and intelligence apparatus. Therefore, any company that works with it, from suppliers of clothing and food to suppliers of compute and AI, are supporting an organization with that mission.
The US military _does not_ need to build autonomous weapon systems and _should not_ surveil US citizens broadly.
I know the reaction to this, if you're a rational observer, is "OpenAI have cut corners or made concessions that Anthropic did not, that's the only thing that makes sense."
However, if you live in the US and pay a passing attention to our idiotic politics, you know this is right out of the Trump playbook. It goes like this:
* Make a negotiation personal
* Emotionally lash out and kill the negotiation
* Complete a worse or similar deal, with a worse or similar party
* Celebrate your worse deal as a better deal
Importantly, you must waste enormous time and resources to secure nothing of substance.
That's why I actually believe that OpenAI will meet the same bar Anthropic did, at least for now. Will they continue to, in the same way Anthropic would have? Seems unlikely, but we'll see.
Another good question: If OpenAI knew Anthropic wasn't a competitor... was the price higher? Will the federal government also pay more for a worse product?
Perhaps Trump's DOD objects specifically to Anthropic models themselves declining to do immoral and illegal things, and not something just stipulated in an ignorable contract. That would give room for Sam to throw some public CYA into a contract, while neutering model safety to their requirements.
How did they agree to the terms that were initially put forward by Anthropic but with OpenAI? Surely there’s a catch here. Or is it just Sam negotiation skill?
They're pretending like they didn't enter into this agreement last January and are completely entrenched in intelligence programs already. They are trying to make it look like they are stepping up in a time of need (time of need for the DoD), in reality they sold their soul to intelligence and the military a year ago.
I posted about this here after Sam made his tweet:
Sam must not be aware of what happened to any business or foreign nation/leader considered outwardly friendly to the first Trump administration when the democrats regained control in 2020.
We really need a plan for the scenario in which the US loses the trade war and decides to go homicidal AI on the whole world. Like, help them recover or something.
At this stage, everything OpenAi does is to try to keep investors investing.
They’re willing to let their brand go to trash for this government contract.
Pretty much every American is standing with Anthropic on this. No one left or right wants mass surveillance and terminators. In fact, no one in the world wants this, except the US military.
But Altman seems so desperate to keep the cash coming he’s ready to do anything.
I have just canceled all services and deleted my account with OpenAI. They can get money from the current US regime but I will not contribute to their violations of the constitution.
So basically Greg Brockman of OpenAI, currently the largest MAGA PAC donor, used his bribe to make the government destroy his main competition? I’m absolutely cancelling ChatGPT and will tell everyone I know to cancel as well.
I also absolutely do not trust sleezy Sam Altman when he claims he has the same exact redlines as Anthropic:
> AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
If Hegseth and Trump attack Anthropic and sign a deal with OpenAI under the same restrictions, it means this is them corrupting free markets by picking which companies win. Maybe it’s at the behest of David Sacks, the corrupt AI czar who complained about lawfare throughout the Biden administration but now cheers on far worse lawfare.
So it’s either a government looking to surveil citizens illegally or a government that is deeply corrupt and is using its power to enrich some people above others.
I hope so but I am less optimistic. The oligarchy in Russia who remained loyal to the Putin regime have done just fine for decades as long as they did not attempt to overthrow the dictator. The regime in Washington is basically constructing the same type of kleptocracy and very little evidence is there that anyone who matters will get in their way. So far as I can tell the country is already a form of authoritarian regime where the loyalty to the supreme ruler is the main parameter of conducting business there.
There's a lot of people in this thread that assume that Sam Altman is the one who is being dishonest here, and I kind of understand, but the other two parties who could just as easily be lying are Pete Hegseth and Donald Trump, and of the three of them if you think sama is the _most_ likely to lie I feel like you have not been paying attention.
So nice of him! I am sure he believes they should offer these terms to all competitors.
HN: if you continue to subscribe to OpenAI, if you use it at your startup, you’re no better than the tech bros you often criticize. This is not surprising but beyond shady.
"Tonight, we reached an agreement with the Department of War to deploy our models in their classified network.
In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome.
AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
We also will build technical safeguards to ensure our models behave as they should, which the DoW also wanted. We will deploy FDEs to help with our models and to ensure their safety, we will deploy on cloud networks only.
We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept. We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements.
We remain committed to serve all of humanity as best we can. The world is a complicated, messy, and sometimes dangerous place."
> We also will build technical safeguards to ensure our models behave as they should
A bold statement. It would appear they've definitively solved prompt injection and all the other ills that LLMs have been susceptible to. And forgot to tell the world about it.
It feels like Hackernews is getting brigaded by Anthropics marketing team. So many posts with people (bots?) talking about cancelling their OAI sub / switching to Claude.
Raise your hand if you actually read it or if you read the title and replied? I see a lot of comments that sure seem like they didn’t read it.
> Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
IF this is true, it SHOULD be verifiable. So, we wait? I mean, I am a dummy, but that language doesn’t seem too washy too me? Either it’s a bold face lie and OpenAI burns because of it or it’s true and the Trump admin is going after the “left” AI company. Or whatever. My point is, someone smarter than me/us is going to fact check Sam’s claim.
The problem is that many of those would-be fact checkers have massive incentives to lie about it. So regardless of whether it is true, you're going to see a number of detailed and well-researched pieces over the weekend arguing that Altman is right and this whole thing is Anthropic's fault. The set of people who could cause OpenAI to burn and the set of people who have millions of dollars riding on its success substantially overlap; it may not take a particularly good argument to convince them.
> Either it’s a bold face lie and OpenAI burns because of it
Do you really still genuinely believe in this? This is the same person that said ads is going to be the last resort, and yet we are getting ads. I just don't understand how people can trust a single word coming out of folks like Sam, Musk, Trump or whoever rich asshole.
I listen to these people talk and they literally do not have souls. They will say whatever it is they need to get ahead. I watched a couple of Sam speeches and videos, the man does not have anything interesting to say.
I like the idea of seeing someone post “I dislike and distrust Sam Altman” and thinking “They must be saying that because they haven’t read the things that he writes”
Do you know who isn't a dummy? Sam. The crucial part of that statement is that the DoD will use OpenAI systems "lawfully and responsibly," which I don't doubt is written somewhere in their contract. However, those terms are so open-ended that it's impossible for OpenAI to enforce. Sam could have clarified in his tweet that they explicitly prohibited the use of their technology for mass surveillance and autonomous killings, but he deliberately chose not to and to simply say, "We told them not to do bad things." which smells like bullshit
No contract can require the government to “reflect” something in law, aside from the fact that the DoD is not a legislative body. So whatever Sam is talking about can only be lip service.
I don't see how OpenAI employees who have signed the We Will Not Be Divided letter can continue their employment there in light of this. Surely if OpenAI had insisted upon the same things that Anthropic had, the government would not have signed this agreement. The only plausible explanation is that there is an understanding that OpenAI will not, in practice, enforce the red lines.
Didn't the safety-conscious employees already leave when OpenAI fired Sam Altman and then re-hired him?
In my mind the only people left are those who are there for the stocks.
In all seriousness, what’s the average tenure at OpenAI and how much of the company in March 2026 was even around for that?
It's comforting to know that some of the brightest minds of our generation are going to work at OpenAI, then quitting a few months later horrified, only to post a short mysterious tweet warning everyone of the dangers ahead. So much for alignment and serving humanity.
Review the signers https://notdivided.org
> Surely if OpenAI had insisted upon the same things that Anthropic had, the government would not have signed this agreement.
But they did.
"Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement."
The difference is that Anthropic wanted to reserve the right to judge when the red lines are crossed, while OpenAI will defer to the DoD and its policies for that. In both cases, the two parties can claim to agree on the principles, but when push comes to shove, who decides on whether the principles are violated differs.
Seems Anthropic did not understand the questions they were asked. From the WaPo:
>A defense official said the Pentagon’s technology chief whittled the debate down to a life-and-death nuclear scenario at a meeting last month: If an intercontinental ballistic missile was launched at the United States, could the military use Anthropic’s Claude AI system to help shoot it down?
>It’s the kind of situation where technological might and speed could be critical to detection and counterstrike, with the time to make a decision measured in minutes and seconds. Anthropic chief executive Dario Amodei’s answer rankled the Pentagon, according to the official, who characterized the CEO’s reply as: You could call us and we’d work it out.
>An Anthropic spokesperson denied Amodei gave that response, calling the account “patently false,” and saying the company has agreed to allow Claude to be used for missile defense. But officials have cited this and another incident involving Claude’s use in the capture of Venezuelan leader Nicolás Maduro as flashpoints in a spiraling standoff between the company and the Pentagon in recent days. The meeting was previously reported by Semafor.
I have a hunch that Anthropic interpreted this question to be on the dimension of authority, when the Pentagon was very likely asking about capability, and they then followed up to clarify that for missile defense they would, I guess, allow an exception. I get the (at times overwhelming) skepticism that people have about these tools and this administration but this is not a reasonable position to hold, even if Anthropic held it accidentally because they initially misunderstood what they were being asked.
https://web.archive.org/web/20260227182412/https://www.washi...
"It’s the kind of situation where technological might and speed could be critical to detection and counterstrike"
Missile detection and decision to make a (nuclear) counterstrike are 2 different things to me but apparently the department of war wants both, so it seems not "just" about missile detection.
> The difference is that Anthropic wanted to reserve the right to judge when the red lines are crossed, while OpenAI will defer to the DoD and its policies for that.
You learned this where?
I’m reading between the lines of the involved parties’ various statements, but there’s also this: https://x.com/UnderSecretaryF/status/2027594072811098230
> I’m reading between the lines of the involved parties’ various statements
You should have said this.
> https://x.com/UnderSecretaryF/status/2027594072811098230
Thank you.
This. Sam is going to pretend they aren’t going to use it for that because his company is collapsing in losses. He will never audit.
Probably also got assurances about a bailout when OpenAI collapses.
I'm sure it's a matter of interpretation. Anthropic thinks the DoW's demands will lead to mass surveillance and auto-kill bots. The DoW probably disagrees with that interpretation, and all OpenAI needs to do is agree with the DoW.
My bet is that what the DoW wants is pretty clearly tied to mass surveillance and kill-bots. Altman is a snake.
Why do you choose to call it the "DoW"? Its official name is the Department of Defense, it was titled that way by Congress and only Congress can change it. What is your motivation in using a term that the current administration has started to use? Do you also use the Gulf of America when referrring to the body of water that defines the southern edge of the USA?
Human responsibility is not the same as human decision making.
And they are crossing the picket line, which honestly I was sure they would do, though I did expect it to take a bit longer.
This is too transparent even for sama.
>Human responsibility is not the same as human decision making.
this is going to end up being interpreted as "well, the president signed off on the operation. see - there's a human in the loop!" - is it?
You wouldn't perchance be in the market for a bridge in the Brooklyn area, would you?
https://news.ycombinator.com/item?id=47190644
Good ole Sammy has never lied
If your starting position is already that Sam Altman lies about everything that doesn't fit your preconceived positions, that doesn't seem like a very useful meaningful position to update.
The company started with a lie, it's in the name.
Unrelated, but want to buy a bridge?
You could recoup your investment in a year by collecting toll. Expedited financing available on good credit!
Please don’t do this here.
Yes, what is implied in this episode is that all big companies that do AI development or provide computing for Ai are now signing for these very shady uses of their technologies.
>Surely if OpenAI had insisted upon the same things that Anthropic had, the government would not have signed this agreement.
Have we been watching the same Trump admin for the last year? That sound exactly like something the government would do: pointlessly throw a fit and end up signing a worse deal after blowing up all political capital.
While that thought crossed my mind, someone in a sub thread of parent comment made a point: OpenAI made a statement about how "We insisted this be not be used in those ways and DoD totally says they won't". Which sounds to me like they ceded any hard terms oand conditions and are letting the DoD use it in "any lawful means" which is what Anthropic didn't stand for.
They seem moderately competent at doing blatant corruption ( https://coinmarketcap.com/currencies/official-trump/ , Qatari jet, etc...). See jeffbee's comment below.
> The only plausible explanation is that there is an understanding that OpenAI will not, in practice, enforce the red lines.
Do you mean the same OpenAI that has a retired U.S. Army General & former director of the NSA (Gen. Nakasone) serving on its board of directors?
All of us can act too. Stop using the OpenAI models. Stop using the app. Design in other models no matter what. Screw these guys.
Do you expect that to work?
No, I expect you to die, Mr. Bond.
Nah. It's possible that the agreement still supports the required terms.
There is more to this story behind the scenes. The government wanted to show power and control over our companies and industries. They didn’t need those terms for any specific utility, they wanted to fight “woke” business that stood up to them.
Supposedly OpenAI had the same terms as Anthropic (according to SamA). Maybe they offered it cheaper and that’s why they agreed. Maybe it’s all the lobbying money from OpenAI that let the government look the other way. Maybe it’s all the PR announcements SamA and Trump do together.
>Supposedly OpenAI had the same terms
"we put them into our agreement." is strange framing is Altman's tweet. Makes me think the agreement does mention the principles, but doesn't state them as binding rules DoD must follow.
None of those explanations are compatible with the pledge of solidarity in the We Will Not Be Divided letter.
> Supposedly OpenAI had the same terms as Anthropic (according to SamA).
He said human responsibility. Anthropic said human in the loop.
And Anthropic refused to say any lawful purpose would be allowed reportedly.
I prescribe literally zero truth value to what Sam says. He will say whatever he needs to get ahead. It is honestly irritating to me that you and many others here seem to implicitly assume his messages are correlated with truth, doing his social engineering work for him, as if his word should adjust your priors even slightly.
I don't necessarily think he's lying, but there's so much obvious incentive for him to lie here (if only because his employees can save face).
Your comment reminded me that a blog post. It’s by the same guy that wrote “programming sucks”. I’ve been sharing it a lot recently lol
https://www.stilldrinking.org/stop-talking-to-technology-exe...
> I don't necessarily think he's lying
He doesn't even need to be lying, the comment is vague and contains enough loopholes that it could be true yet meaningless. I explained some that I noticed here: https://news.ycombinator.com/item?id=47190163
I'm not sure if I'd go down to zero, but he did get fired from OpenAI for lying.
And fired from YC for lying. And lied to investors about how many Loopt employees he had. And lied about having 100x the actual number of users when he sold it.
It's this simple: Trump is a criminal. Larry Ellison is his pal. Sam Altman has a huge deal for cloud services from Oracle. Trump is using the DoD budget to backstop Ellison's business.
This is pretty much on the right take on it, although it's much more than that. It's very clear at this point, especially the first conclusion, but people insist in looking to the other side.
Interesting thesis.
But regardless of the moral implications, will this improve America’s position on the global stage or further undermine it?
Only if you think that crime will somehow improve America. My opinion is that this is leading to its collapse, no matter how "powerful" they look.
Attempting to kneecap the breakout front runner of the major American AI companies to ensure the shittier, politically compliant one wins in the short term? Gee I wonder.
Anthropic is great but not the undisputed front runner.
I can also interpret this as Sam and the administration supporting accelerationism while Dario is more measured and wishes to slow things down.
For better or worse, outright nationalization of military related companies is common on a global scale. I plan to do my best to ensure this is a domestic catastrophe, and I hope we'll succeed, but I don't expect other countries to care much about varying levels of regime alignment between two billionaire American defense contractors.
Maybe Sam Altman said nicer things about Donald Trump. Maybe he promised that he would not revoke their API keys when Hegseth directs the military to seize ballots. Maybe he's jockeying for position to take over the government when AGI hits.
Ultimately, I don't know how much the specific reasons matter. Pete Hegseth must be removed from office, OpenAI must be destroyed for their betrayal of the US public, that's all there is to it.
1) Another OpenAI cofounder (Brockman) gave Trump’s superPAC the largest ever individual donation of $25m.
2) Trump’s son in law (Kushner) has most of his net worth wrapped up in OpenAI.
don't forget that Sama is a Thiel protege
If you're unhappy with this, an immediate way to signal it is with your wallet. In my case I've just uninstalled chatgpt from my phone, cancelled my subscription and will up my spend with anthropic.
Thanks for the reminder. Doing the same now.
The little respect I had left for Sam is now wiped. Makes me sick.
Growing up I always thought AI would be this beautiful tool, this thing that opens the gates to a new society where work becomes optional in a way. But I failed to think about human greed.
I remember following OpenAI way back when it was a non profit explaining how AI uncontrolled could be highly detrimental. Now Sam has not only taken that non profit and made it for-profit. It seems he’s making the most evil decisions he can for a buck.
Cancel your subscription, tell your friends to. And vote to heavily tax these companies and their leaders.
Perfect timing - Had already cancelled my Claude sub over their OAuth ban in external tools and was about to pick up a Codex sub as the next best alternative.
Ended up renewing my Claude sub today instead. Principled stances matter and I no longer trust OpenAI to be trustworthy custodians of my AI History.
Just canceled my subscription! I immediately received an email with the subject “We’d love your feedback on why you canceled your ChatGPT plus subscription” and a link to a survey.
I linked to https://notdivided.org/ as the reasoning why.
I’d like to say I did that but I already canceled my subscription 4 months ago in favor of Claude and Gemini based purely on product quality.
Was shocking back then to think how far we’ve come.
Deleted all chats and deleted my account.
I tried doing that but I'm certain they didn't delete it, because I tried logging in after a week and it worked
Totally agree. Signed up for a claude code account and will not give OpenAI any money in the future. Let's see what Google does. I will definitely vote with my wallet.
Just deleted OpenAI account, F these guys
Thanks for reminding me. Been meaning to cancel for months.
Exactly. Stop using OpenAI. Don’t design it into your software at work. Use Claude. Screw these guys.
Same
I canceled my subscription, wiped my history, closed my account, deleted the app. Using Claude Max.
Personally I'm happy about this. OpenAI are being fair about letting the gov use their models to spy on everybody, doesn't seem right that Americans get a pass.
Do you honestly believe that cancelling a subscription makes a bit of difference to a company that is either committing accounting fraud on a monumental scale or shoveling venture capital money into a furnace? not to mention the whole collaborating with a fascist government thing.
taking real action is your choice, but stop pretending this kind of thing matters one iota
edit: to be clear, i'm not advocating for nihilism, but tricking yourself into thinking you made a difference to make yourself feel better isn't the play either
It absolutely matters, especially when done in unison like this.
Cancelling ChatGPT sends a signal that you don't agree with weaponizing AI. Switching to Claude says you support Anthropic's principled stance against it. If you have a strong opinion either way, today is the day to vote with your wallet.
Dismissing every small action as meaningless is just apathy and how nothing ever changes.
Anthropic isn't against weaponizing AI, it's just against two specific carve outs for now. They happily accepted the Pentagon's money so long as it was only spying on other countries. And now that the leopard is eating their face they're claiming the moral high ground.
It's entirely possible for both Anthropic and OpenAI to be in the wrong here. This is a massive publicity win but it doesn't make them heroes in my book.
What has an impact is cancelling a subscription and then talking about it. The media will amplify it the pushback. The goal is to make the name OpenAI and ChatGPT toxic, that whatever you do will be converted into a technology that will surveil or bomb you.
At least I'm not getting my hands dirty.
Yes? Earnings matter to investors
Do they? What are those OpenAI earnings that you are talking about? That's a company that should have ceased existing some time ago if earnings were important
Investors want to see growth. If there’s no growth or even a loss in users the next round of funding will be more difficult to secure.
> but stop pretending this kind of thing matters one iota
This is blatantly false and intellectually dishonest. Of course it matters. Your edit is also wrong; you are advocating for nihilism with statments like these.
It's the only thing that matters. These companies don't follow the rules of capitalism physics. They live or die on vibes alone and the tech community abandoning them en masse is bad for the vibes. Once they lose the vibes they are Wiley Coyote looking down at the canyon below.
I think you have too much pessimism. It's not guaranteed to work, but as I mentioned in another thread, since around December, Claude (and Gemini to a lesser extent) has had all the buzz in tech circles, while Chat-GPT has seemed like the also-ran. And that matters: decision-makers in companies notice these things and momentum becomes self-reinforcing (you use Claude Code because everyone else uses Claude Code). If a large enough group of developers visibly defects from OpenAI because of this, it definitely could have consequences. It's not a sure thing, but it's far from hopeless.
I was not a Chat-GPT user even before this, but I'm bumping my Claude Code subscription to the next tier up. Fuck OpenAI.
My knee-jerk reaction to this was looks like an opportunistic maneuver that Sam is known for and I'm considering canceling my subscriptions and business with OpenAI
But what's the most charitable / objective interpretation of this?
For example - https://x.com/UnderSecretaryF/status/2027594072811098230
Does it suggest that determination of "lawful use" and Dario's concerns falls upon the government, not the AI provider?
Other folks have claimed that Anthropic planned to burn the contentious redlines into Claude's constitution.
Update: I have cancelled my subscriptions until OpenAI clarifies the situation. From an alignment perspective Anthropic's stand seems like the correct long-term approach. And at least some AI researchers appear to agree.
As people have repeatedly mentioned, if the War Department was unhappy with Anthropic's terms, they could have refused to sign the contract. But they didn't: they were fine with it for over a year. And if they changed their mind, they could've ended the contract and both sides could've walked away. Anthropic said that would've been fine. But that's not what happened either: they threatened Anthropic with both SCR designation and a DPA takeover if Anthropic didn't agree to unilateral renegotiation of terms that the War Department had already agreed were fine.
It's absurd, and doubly so if OAI's deal includes the same or even similar redlines to what Anthropic had.
Sam is just about the least trustworthy person in AI, I don't trust his words as face value and I consider these weasel words:
> prohibitions on domestic mass surveillance and human responsibility *for the use of force*
That means autonomous killbots are a-ok. Human responsibility is not the same as human decision-making.
The president or anybody at DoD can be "responsible", and we know there will be zero accountability. The courts defer to the executive, and Congress is all-too-happy for the executive to take the flak for their wars.
>human responsibility for the use of force, including for autonomous weapon systems
So there’s the difference, and an erasure of a red line. OpenAI is good with autonomous weapon systems. Requiring human responsibility isn’t saying much. Theres already military courts, rules of engagement, and international rules of war.
More details on the difference between the OpenAI and Anthropic contracts from one of the Under Secretaries of State:
>The axios article doesn’t have much detail and this is DoW’s decision, not mine. But if the contract defines the guardrails with reference to legal constraints (e.g. mass surveillance in contravention of specific authorities) rather than based on the purely subjective conditions included in Anthropic’s TOS, then yes. This, btw, was a compromise offered to—and rejected by—Anthropic.
https://x.com/UnderSecretaryF/status/2027566426970530135
> For the avoidance of doubt, the OpenAI - @DeptofWar contract flows from the touchstone of “all lawful use” that DoW has rightfully insisted upon & xAI agreed to. But as Sam explained, it references certain existing legal authorities and includes certain mutually agreed upon safety mechanisms. This, again, is a compromise that Anthropic was offered, and rejected.
> Even if the substantive issues are the same there is a huge difference between (1) memorializing specific safety concerns by reference to particular legal and policy authorities, which are products of our constitutional and political system, and (2) insisting upon a set of prudential constraints subject to the interpretation of a private company and CEO. As we have been saying, the question is fundamental—who decides these weighty questions? Approach (1), accepted by OAI, references laws and thus appropriately vests those questions in our democratic system. Approach (2) unacceptably vests those questions in a single unaccountable CEO who would usurp sovereign control of our most sensitive systems.
> It is a great day for both America’s national security and AI leadership that two of our leading labs, OAI and xAI have reached the patriotic and correct answer here
https://x.com/UnderSecretaryF/status/2027594072811098230
A government promise that they'll only do lawful things is not reassuring at all:
1. We've seen government lawyers write memos explaining why such-and-such obviously illegal act is legal (see: torture memo). Until challenged, this is basically law.
2. We've seen government change the law to make whatever they want legal (see: patriot act)
3. We've seen courts just interpret laws to make things legal
A contractor doesn't realistically have the power to push back against any of these avenues if they agree to allow anything legal.
(At the risk of triggering Godwin's Law, remember that for the most part the Holocaust was entirely legal - the Nazi's established the necessary authorization. Just to illustrate that when it comes to certain government crimes, the law alone is an insufficient shield.)
You're quoting social media posts from a regime official who says he didn't participate in these negotiations and doesn't work for the relevant department.
If his characterization of the agreement is correct, which I will not believe and you should not believe until a trustworthy news outlet publishes the text, I suppose this would convince me that Hegseth does not literally plan to build a Terminator for democracy-ending purposes. There's a lot of inexcusable stuff here regardless, but perhaps merely boycotting OpenAI and the US military would be a sufficient response if this all checks out.
> which I will not believe and you should not believe
It seems like you chose to immediately disbelieve it.
> until a trustworthy news outlet publishes the text
If you've found one of these, let me know. I'm still looking...
[delayed]
If the redlines are the same how'd this deal get struck?
ChatGPT maker OpenAI has the same redlines as Anthropic when it comes to working with the Pentagon, an OpenAI spokesperson confirmed to CNN.
https://edition.cnn.com/2026/02/27/tech/openai-has-same-redl...
You're expecting logic from the Trump administration and that's not really how they do things. Maybe it was never about the redlines? Maybe they decided Anthropic was their enemy, and that was their excuse.
Anthropic was too public about being “good”. And if there is one thing the Trump admin cannot abide it’s morality.
Just uninstalled the app and canceled subscription. OpenAI can't justify their insane valuation without an user base. Especially when there are capable models elsewhere.
So they agreed to the same red lines that had earlier led to the fallout with Anthropic? Kind of strange.
I bet Sam secretly pledged to DoD that the red lines were only temporary, for optics and to calm employees at the all hands meeting.
A few months down the line, OpenAI will quietly decide that their next model is safe enough for autonomous weapons, and remove their safeguard layer. The mass surveillance enablement might be an indirect deal through Palantir.
Very possible, double speaking is Sam Altman's specialty.
Sam saw Anthropic was getting too competitive. So he called his buddies in the gov to knock them down a peg.
That's very possible! In the last few days Anthropic was getting a lot of attention, and OpenAI was looking weaker in comparison. It seems like a politically coordinated job to remove competition.
For sure, he's been pissed that OpenAI no longer has the Mandate of Heaven and Claude is all anyone has been talking about since December. (And it's not just an ego thing: because OAI isn't profitable yet, they need the hype to keep going to raise money on favorable terms, so loss of buzz is an existential threat). I absolutely believe that he started making calls to try and get buddies in the White House to take Anthropic down.
I don't trust Sam to be telling the truth. It would be to his benefit to lie about this and make Anthropic look bad, so he of course would, even if it's not actually the case.
Well you know how it goes... you need to read between the lines. I can agree with you on your "principles", but not enforce them myself.
It makes sense if you imagine the real motivation is “make sure the AI contracts go to my good friend Sam”, and all the red line stuff is just a way to pick a fight with Anthropic.
No, the difference is that the government agrees to no "unlawful" use as determined by the government.
Anthropic said that mass surveillance was per se prohibited even if the government self-certified that it was lawful.
Follow the money. There is a UAE sheik who bought 49% of Trump's World Liberty and is involved in OpenAI's Project Stargate:
https://www.binance.com/en/square/post/35909013656801
I'm sure more will drop in the coming months.
I had kept my Plus subscription just because I was lazy, and it was inexpensive and convenient… but this turn definitely helped me get off the fence. I am exporting and deleting my data now, and the cancellation is already done.
We need some kind of group like "tech people with morals". I'm done with these people and their corruption and garbage.
Yeah some new banner to organise around- the hard part is easily communicating you're an ethical technologist and finding others.
"Starting today I will be asking prominent members of the tech community to sign their name onto this. A code of conduct, authored by me, that pledges them to a universal ethos, which I created, that I call tech ethics or Tethics for short."
Also, it's probably tricky to find a Schelling point that a broad range of people can agree to.
* no military use
* no lethal use
* no use in support of law enforcement
* no use in support of immigration enforcement
* no use in mass surveillance
* no use in domestic mass surveillance (but mass surveillance of foreigners is OK)
* no use in domestic surveillance
* no use in surveillance
* require independent audits
* require court oversight
* require company to monitor use
* require company to monitor use and divulge it to employees
* some other form of human rights monitoring or auditing
* some other form of restriction on theaters/conflicts/targets
* company will permit some of these uses (not purport to forbid them by license, contract, or ToS) but not customize software to facilitate them
* company can unilaterally block inappropriate uses
* company can publicly disclose uses it thinks are inappropriate
* some other form of remedy
* government literally has to explain why some uses are necessary or appropriate to reassure people developing capabilities, and they have some kind of ongoing bargaining power to push back
It feels normal to me that a lot of people would want some of those things, but kind of unlikely that they would readily agree on exactly which ones.
I even think there's a different intuition about the baseline because one version is "nobody works on weapons except for people who specifically make a decision to work for an arms company because they have decided that's OK according to their moral views" (working on weapons is an abnormal, deliberate decision) and another version is "every company might sell every technology as part of a weapons system or military application, and a few people then object because they've decided that's not OK according to their moral views" (refusing to work on weapons is an abnormal, deliberate decision). I imagine a fair number of people in computing fields effectively thought that the norm or default for their industry was the latter, because of the perception that there are "special" military contractors where people get security clearances and navigate military procurement processes, and most companies are not like that, so you were not working on any form of weapon unless you intentionally chose to do so. But, having just been to the Computer History Museum earlier this week, I also see that a lot of Silicon Valley companies have actually been making weapons systems for as long as there has been a Silicon Valley.
There is definitely a muddle on so many levels about signaling and agreeing on ethics in technology.
But as innovation slows globally, it is implementation, ethics, and ideology that will once again be the dominant metrics of progress, so there's a new window emerging to push for this social/moral change in technology once again.
So it's still critically important that we actively work towards finding a meaningful, socially contagious differentiator other than "ethical technologist" even if it's difficult- look at what OpenAI gets away with under that flimsy banner.
I would put bets on the issue probably being that it was pointed out that Anthropic's models were used to assist the raid in Venezuela, Anthropic then aggressively doubled down on their rules/principles and the DOD didn't like being called out on that so they lashed out, hard.
If theres anything this admin doesn't like, its being postured against or called out by literally anyone, especially in public.
I don't even think Anthropic balked at being used to assist, as long as a human has the final say.
Cancel your subscription. It's the least you can do.
All OpenAI employees during the board revolt that vouched for sama's return are personally responsible.
OpenAI employees revolted for their millions worth of stock, not for principle.
Anyone thinking they have any virtue is naive.
Choosing to go along with calling it the "Department of War" tells you all you need to know.
I'm unsure how to feel about this whole dust-up. It doesn't seem like much has changed in substance. Maybe OpenAI outmaneuvered Anthropic behind the scenes. Possibly Anthropic was seen as not behaving deferentially enough towards the government. But this administration has proven comically corrupt, so it wouldn't surprise me if money was involved. Will be interested to see what journalists turn up.
It’s amazing how quickly the players keep shifting here.
Yesterday and the day before sentiment seemed to be focused on “Anthropic selling out”, then that shifted to “Anthropic holds true to its principles in a David vs Goliath” and “the industry will rally around one another for the greater good.” But suddenly we’re seeing a new narrative of “Evil OpenAI swoops in to make a deal with the devil.”
Reminds of that weekend where Sam Altman lost control of OpenAI.
"There are decades where nothing happens and weeks where decades happen."
Sam is a player and honestly the more interesting one in the whole thing.
Mad respect to Sam, now I believe OpenAI have better chance to win in the race
> Mad respect to Sam
And people wonder how we got here.
He’s certainly solidified his place in the history of this era.
But I suspect the public sentiment will eventually turn against him. When society sets its pitchforks on big tech he’ll be the poster boy. A 21st century John D. Rockefeller.
Him, Musk, Bezos, and Zuck.
Hitler won the race in the 1930s too. Totally crushed it.
I considered that comparison, but in all seriousness, I’m not sure it’s apt.
Are he and his peers Hitler or they the naive oligarchs who think they can keep populist leaders and their constituencies under their thumb? Only to be out maneuvered by the people who the masses think have their back.
I know many folks who think their political leaders have the best interest at heart (rightly or wrongly). I know nobody who thinks tech leaders do. At best they want to be them.
This is awkward? https://news.ycombinator.com/item?id=47188473
So they agreed to the exact same clauses that Anthropic put forward but with OpenAI instead?
So it wasn't about those principles making them a supply chain risk? They're just trying to punish Anthropic for being the first ones to stand firm on those principles?
I’m sure a big donor just used the US gov as a bludgeon to destroy their competition
Is the big donor among us?
Now that OpenAI is going to be used for mass domestic surveillance you can assume Sam Altman is always in the room.
As I understand it, Sam's cofounder at OpenAI donated $25 million to the Trump 2024 campaign.
As Trump himself likes to say, "Promises made, promises kept."
Anthropic would probably not renegotiate in a year about the principles, while Sam Altman is known to be morally flexible so OpenAI will almost surely allow the military to do what they want in the future. Sam Altman might even have said behind closed doors that these restrictions will be removed once the drama has died down.
Absolute disgrace of a person and organization.
Is the Pentagon signing a EULA confirming all their data will now be used, anonymised, for improving the service?
Obviously not? You know enterprise customers don't have the same EULA as consumers, right?
For the people that don't understand how they got a deal with the same redlines, it probably because OpenAI agreed to not question them. The safeguards are there, both parties agree now fuck off and let us use your model how we see fit.
Anthropic probably made the mistake of questioning the Military's activities related to Claude after the Venezuela mission and wanted reassurance that the model wouldn't be used for the redlines, and the military didn't like this and told them we aren't using your models unless you agree to not question us and then the back and forth started.
In the end, we will probably have both OpenAI and Anthropic providing AI to the military and that's a good thing. I don't think they will keep the supply chain risk on Anthropic for more than a week.
Anthropic vs OpenAI will probably be The Machine vs Samaritan
(Person Of Interest for those who haven't seen it, watched it a decade ago and it's actually quite surprising how on point it ended up being)
cancelling my openai subscription, they're gonna miss my 20 USD
Did anyone ever doubt sama would just follow the money?
weasels gonna weasel
China has evacuated its embassies in Iran.
This is really about the imminent strike on Iran which is now super telegraphed. They are gonna use ChatGPT for target selection, and the likely outcome is that it will fuck things up and a bunch of civilians are going to die because of this decision.
When this happens, Altman will go from being merely a drifter to having blood on his hands.
So does this mean that OpenAI will give whatever the DoD asks for and they will pinky swear that it won’t be used for mass surveillance and autonomous killing machines?
yes
and we know we can trust openAI because they were founded on "open" and "safe" AI (up until they realized how much money there was to be made, at which point their only value changed to "make money")
I'm never using an OpenAI model or Codex ever again. Period. Idaf whether it scores better than Claude on benchmarks or not.
This is a red line for me. It's clear OpenAI has zero values and will give Hesgeth whatever he wants in exchange for $$$.
https://www.nytimes.com/2026/02/27/technology/openai-reaches...
Sam Altman is this. Sam Altman needs to be stopped.
Screw OpenAI. Never opening that app again or using one of their models.
What a snake
This seems full of loopholes.
> The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
(1) Well, did both sides sign the agreement and is it actually effective? Or is it still sitting on someone's desk until it can get stalled long enough?
(2) What does "agreement" even mean? Is it a legally enforceable contract, or just some sort of MoU or pinkie promise?
(3) If it's a legally enforceable contract, is it equally enforceable on all of their contracts, or just some? Do they not have existing contracts this would need to apply to?
(4) What does "reflects them in law and policy" even mean? Since when does DoW make laws, and in what sense do their laws reflect whatever the agreement was? Are these laws he can point to so everyone else can see? Can he at least copy-paste the exact sentences the government agreed to?
Instant uninstall.
All that money and not a single ounce of integrity.
So this is indeed how OpenAI survives (a little bit longer ?) - government bailout.
Ah, is it the time when Skynet starts to manifest itself...
So there are two possibilities here:
1. There's no substantive change. Hegseth/Trump just wanted to punish Anthropic for standing up to them, even if it didn't get them anything else today -- establishing a chilling effect for the future has some value for them in this case, after all. And OpenAI was willing to help them do that, despite earlier claiming that they stood behind Anthropic's decisions.
2. There is a substantive change. Despite Altman's words, they have a tacit understanding that OpenAI won't really enforce those terms, or that they'll allow them to be modified some time in the future when attention has moved on elsewhere.
Either way, it makes Altman look slimy, and OpenAI has aligned with Trump against Anthropic in a place where Anthropic made a correct principled stand. It's been clear for a while that Anthropic has more ethics than OpenAI, but this is more naked than any previous example.
> OpenAI has aligned with Trump against Anthropic in a place where Anthropic made a correct principled stand.
Just to be clear, you believe that the correct, principled stand is that it's OK to use their models for killing people and civilian surveillance?
Both OAI and Anthropic have the same moral leg to stand on here, OAI is just not hypocritical about it.
If you believe that any country should have a military and intelligence apparatus, the job of that apparatus is to kill people and surveil foreigners. I do think the US government should have a military and intelligence apparatus. Therefore, any company that works with it, from suppliers of clothing and food to suppliers of compute and AI, are supporting an organization with that mission.
The US military _does not_ need to build autonomous weapon systems and _should not_ surveil US citizens broadly.
I know the reaction to this, if you're a rational observer, is "OpenAI have cut corners or made concessions that Anthropic did not, that's the only thing that makes sense."
However, if you live in the US and pay a passing attention to our idiotic politics, you know this is right out of the Trump playbook. It goes like this:
* Make a negotiation personal
* Emotionally lash out and kill the negotiation
* Complete a worse or similar deal, with a worse or similar party
* Celebrate your worse deal as a better deal
Importantly, you must waste enormous time and resources to secure nothing of substance.
That's why I actually believe that OpenAI will meet the same bar Anthropic did, at least for now. Will they continue to, in the same way Anthropic would have? Seems unlikely, but we'll see.
Another good question: If OpenAI knew Anthropic wasn't a competitor... was the price higher? Will the federal government also pay more for a worse product?
You'd have to think so. They're really the only serious player left - I doubt Google would want to be involved, and xAI is a significant step down.
Not a surprise here, that letter was a trap for OpenAI employees who filled it out with their names on it. [0]
The ones that did might as well leave. But there was no open letter when the first military contract was signed. [1] Now there is one?
[0] https://news.ycombinator.com/item?id=47176170
[1] https://www.theguardian.com/technology/2025/jun/17/openai-mi...
Perhaps Trump's DOD objects specifically to Anthropic models themselves declining to do immoral and illegal things, and not something just stipulated in an ignorable contract. That would give room for Sam to throw some public CYA into a contract, while neutering model safety to their requirements.
How did they agree to the terms that were initially put forward by Anthropic but with OpenAI? Surely there’s a catch here. Or is it just Sam negotiation skill?
They're pretending like they didn't enter into this agreement last January and are completely entrenched in intelligence programs already. They are trying to make it look like they are stepping up in a time of need (time of need for the DoD), in reality they sold their soul to intelligence and the military a year ago.
I posted about this here after Sam made his tweet:
https://news.ycombinator.com/item?id=47189756
Source: https://defensescoop.com/2025/01/16/openais-gpt-4o-gets-gree...
Snakes- as predicted
Sam must not be aware of what happened to any business or foreign nation/leader considered outwardly friendly to the first Trump administration when the democrats regained control in 2020.
If they earnestly believe in fast ASI timelines then political grudges have to be pretty low on OAI's list of worries about 2029.
You’re assuming democrats will ever be allowed to regain control.
Opportunism without principles at its finest.
I wonder if this will cause this to save open ai from the bubble! i am sure i am wrong;-)
We really need a plan for the scenario in which the US loses the trade war and decides to go homicidal AI on the whole world. Like, help them recover or something.
At this stage, everything OpenAi does is to try to keep investors investing.
They’re willing to let their brand go to trash for this government contract.
Pretty much every American is standing with Anthropic on this. No one left or right wants mass surveillance and terminators. In fact, no one in the world wants this, except the US military.
But Altman seems so desperate to keep the cash coming he’s ready to do anything.
Department of War just killed OpenAI's brand
I have just canceled all services and deleted my account with OpenAI. They can get money from the current US regime but I will not contribute to their violations of the constitution.
perhaps us mere mortals should petition our lawmakers to ban mass surveillance.
So basically Greg Brockman of OpenAI, currently the largest MAGA PAC donor, used his bribe to make the government destroy his main competition? I’m absolutely cancelling ChatGPT and will tell everyone I know to cancel as well.
I also absolutely do not trust sleezy Sam Altman when he claims he has the same exact redlines as Anthropic:
> AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
If Hegseth and Trump attack Anthropic and sign a deal with OpenAI under the same restrictions, it means this is them corrupting free markets by picking which companies win. Maybe it’s at the behest of David Sacks, the corrupt AI czar who complained about lawfare throughout the Biden administration but now cheers on far worse lawfare.
So it’s either a government looking to surveil citizens illegally or a government that is deeply corrupt and is using its power to enrich some people above others.
This will backfire on Sam someday, he’s just a pawn in the agenda of the Trump admin.
I hope so but I am less optimistic. The oligarchy in Russia who remained loyal to the Putin regime have done just fine for decades as long as they did not attempt to overthrow the dictator. The regime in Washington is basically constructing the same type of kleptocracy and very little evidence is there that anyone who matters will get in their way. So far as I can tell the country is already a form of authoritarian regime where the loyalty to the supreme ruler is the main parameter of conducting business there.
There's a lot of people in this thread that assume that Sam Altman is the one who is being dishonest here, and I kind of understand, but the other two parties who could just as easily be lying are Pete Hegseth and Donald Trump, and of the three of them if you think sama is the _most_ likely to lie I feel like you have not been paying attention.
SA is a real weasel lol. Acted like he stood behind Anthropic's principles just to announce the deal with DoW a few hours later.
Sam Altman not being consistently candid or truthful would be the shock of the century.
So nice of him! I am sure he believes they should offer these terms to all competitors.
HN: if you continue to subscribe to OpenAI, if you use it at your startup, you’re no better than the tech bros you often criticize. This is not surprising but beyond shady.
"Tonight, we reached an agreement with the Department of War to deploy our models in their classified network.
In all of our interactions, the DoW displayed a deep respect for safety and a desire to partner to achieve the best possible outcome.
AI safety and wide distribution of benefits are the core of our mission. Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
We also will build technical safeguards to ensure our models behave as they should, which the DoW also wanted. We will deploy FDEs to help with our models and to ensure their safety, we will deploy on cloud networks only.
We are asking the DoW to offer these same terms to all AI companies, which in our opinion we think everyone should be willing to accept. We have expressed our strong desire to see things de-escalate away from legal and governmental actions and towards reasonable agreements.
We remain committed to serve all of humanity as best we can. The world is a complicated, messy, and sometimes dangerous place."
> We also will build technical safeguards to ensure our models behave as they should
A bold statement. It would appear they've definitively solved prompt injection and all the other ills that LLMs have been susceptible to. And forgot to tell the world about it.
/s
Hey dang I know I’m not allowed to say this due to community guidelines, but Sam Altman is a lying sack of shit.
It feels like Hackernews is getting brigaded by Anthropics marketing team. So many posts with people (bots?) talking about cancelling their OAI sub / switching to Claude.
Cancelling seems like a pretty reasonable thing for a human to do.
Raise your hand if you actually read it or if you read the title and replied? I see a lot of comments that sure seem like they didn’t read it.
> Two of our most important safety principles are prohibitions on domestic mass surveillance and human responsibility for the use of force, including for autonomous weapon systems. The DoW agrees with these principles, reflects them in law and policy, and we put them into our agreement.
IF this is true, it SHOULD be verifiable. So, we wait? I mean, I am a dummy, but that language doesn’t seem too washy too me? Either it’s a bold face lie and OpenAI burns because of it or it’s true and the Trump admin is going after the “left” AI company. Or whatever. My point is, someone smarter than me/us is going to fact check Sam’s claim.
1-800-Come-on-now
DoW: WOKE Antropic tried to impose their 'values' on us? Friendship ended!! National security risk!
OpenAI: We just signed a deal that's strong on values, the exact same ones as Anthropic, no way we would mislead anyone about this
You: Seems legit
The problem is that many of those would-be fact checkers have massive incentives to lie about it. So regardless of whether it is true, you're going to see a number of detailed and well-researched pieces over the weekend arguing that Altman is right and this whole thing is Anthropic's fault. The set of people who could cause OpenAI to burn and the set of people who have millions of dollars riding on its success substantially overlap; it may not take a particularly good argument to convince them.
Yeah, you’re right. I’m overly hopeful and naïve
Edit: as soon as I hit submit I realized this might sound condescending, but I actually mean this lol
FWIW I thought the intent was clear.
> Either it’s a bold face lie and OpenAI burns because of it
Do you really still genuinely believe in this? This is the same person that said ads is going to be the last resort, and yet we are getting ads. I just don't understand how people can trust a single word coming out of folks like Sam, Musk, Trump or whoever rich asshole.
I listen to these people talk and they literally do not have souls. They will say whatever it is they need to get ahead. I watched a couple of Sam speeches and videos, the man does not have anything interesting to say.
I like the idea of seeing someone post “I dislike and distrust Sam Altman” and thinking “They must be saying that because they haven’t read the things that he writes”
Do you know who isn't a dummy? Sam. The crucial part of that statement is that the DoD will use OpenAI systems "lawfully and responsibly," which I don't doubt is written somewhere in their contract. However, those terms are so open-ended that it's impossible for OpenAI to enforce. Sam could have clarified in his tweet that they explicitly prohibited the use of their technology for mass surveillance and autonomous killings, but he deliberately chose not to and to simply say, "We told them not to do bad things." which smells like bullshit
I guess I’m hanging on what
> reflects them in law
Means exactly. What law and what does it say?
I’m also sure he quietly bent the knee, but I want to know what “law and policy” it’s being reflected in to know.
No contract can require the government to “reflect” something in law, aside from the fact that the DoD is not a legislative body. So whatever Sam is talking about can only be lip service.
In my experience ChatGPT is the most sanctimonious of the leading models.
When I need advice for my clandestine operations I always reach for Grok.