Is regulation unregulated in the US? As in it’s missing consistency and transparency. Beyond Anthropic, the US signals to the world that depending and using American solutions is high risk. Is that what the US government desires? The digital battleground across regions is on the rise. I don’t see us as the US on the right path.
tomjakubowski 1 days ago [-]
In the US regulation is, in theory, regulated by Congress, which passes laws granting regulation powers to the federal agencies. Congress and the laws it passes are, in theory, regulated by the Constitution, and interpretation thereof by the Supreme Court.
quantified 21 hours ago [-]
Regulation is always by motivation. Charges can be delayed, cops can decide not to arrest, prosecutors can decide not to prosecute, etc.
Government was supposed to be high-trust, but Nixon weakened the trust, Reagan explicitly took aim at it and lots of people went along. Now the reactionaries run things. Cancer in the brain.
sp4cec0wb0y 22 hours ago [-]
And the supreme court is unregulated (except for appointment and impeachment).
ckozlowski 22 hours ago [-]
Well, that and, by passing additional legislation.
There's two parts to this. The first is that in most cases, the issues stem from lack of clarity in existing legislation, or contradictions thereof. At anytime, if Congress does not like the way the Supreme Court ruled, they can pass different legislation.
The second stems in part from the first. If what Congress passed is interpreted as being unconstitutional by the Supreme Court, the States can amend the constitution.
There is no case where the Supreme Court can (in principle) simply make their own rules*
The debate and controversy comes where the Supreme Court is seen as making their own rules (or arguably asked to) because the legislative process is deemed too cumbersome or disadvantageous to a party. In my opinion, this is where a lot of the difficulties lie, in that the Supreme Court is asked to rule on matters that really should be more clearly legislated. But the issues are seen as easier/quicker to address by convincing 9 justices rather than two legislative bodies and a chief executive. Naturally, this brings it's own consequences.
ceejayoz 22 hours ago [-]
> Well, that and, by passing additional legislation.
The Voting Rights Act might disagree on that.
> There is no case where the Supreme Court can (in principle) simply make their own rules.
US Constitution Article III Section 2 Clause 2 which says the supreme court's appellate jurisdiction is subject to "Exceptions, and under such Regulations as the Congress shall make." So Congress could, for example, require the supreme court to hear every case appealed to them instead of letting the justices pick their cases. Or it could require judges to recuse themselves if there is a conflict of interest*. Or it could forbid the court hearing cases where an en banc appeals court was unanimous.
* which could have more teeth than the self-imposed ethics rules for conflict of interest that the court has now.
matheusmoreira 23 hours ago [-]
> Is that what the US government desires?
The US desires to show its own citizens and the world how painful it is not to submit to the US government. They're leveraging everything they've got.
The only rational response when faced with this sort of behavior is to reduce dependence on and exposure to the US as much as possible. Develop national technology. Dedollarize the economy. Invest in and use open weight models.
jstummbillig 22 hours ago [-]
> Is regulation unregulated in the US? As in it’s missing consistency and transparency.
Institutions are not built around zero trust. Because when we assume attackers are identifiable, and there is some level of integrity, that makes things a lot simpler and cheaper.
I will not expect my pediatrician to want to kill my child and they certainly could, but it would be completely bonkers to try to guard against that.
The level of damage that a single really motivated, disgruntled person with a lot of power can do in most settings is intense. The cost inflicted if we tried to build guardrails against all of it is insane, and it probably would never work.
vitally3643 21 hours ago [-]
Actually, the traditional guardrails we've always had are plenty. If a doctor starts trying to kill patients they are removed from society (they go to jail for a long time).
Traditionally this was applied to people in power with the power of a good old fashioned lynch mob. If you obtain great power and use it to start hurting large numbers of people, the citizenry at large would drag that person out into the street and behead or beat them to death.
It worked pretty well for all societies throughout history, including the West. If a person becomes a danger to society, they should be removed from society with proportional prejudice. That's the guardrail society has always had.
Problem is now that those in power have managed to convince large fractions of the population that it's wrong to say mean things about sociopathic mass murderers and child rapists or the guy who is actively trying to kill you personally.
ryandrake 21 hours ago [-]
> If a doctor starts trying to kill patients they are removed from society (they go to jail for a long time).
Unfortunately we don't seem to have the ability to do the same for corporations. If a person starts killing or robbing people, they get arrested and are removed from society. If a corporation starts killing or robbing people, they maybe get a strongly worded letter, maybe have their CEO say a few words in front of Congress, and maybe get a token fine amounting to 0.001% of their revenue (which they will appeal for ten years). But for whatever reason, we don't seem to have the guts to remove a badly behaving corporation from society--no matter what heinous things they are doing, no matter what laws they are breaking.
cadamsdotcom 21 hours ago [-]
> Problem is now that those in power have managed to convince large fractions of the population that it's wrong to say mean things about sociopathic mass murderers and child rapists or the guy who is actively trying to kill you personally.
Actually, civil societies granted a monopoly on violence to the government which vests that in police and military. It’s a far better system than random justice because it’s predictable and you know upfront what’ll trigger it.
That system had its flaws, but despite that it was working pretty well until those flaws were found and exploited allowing it to be captured.
ajju 23 hours ago [-]
The current administration's actions are for sure signaling a distinct lack of predictability or stability.
The U.S. is either harming itself erratically, or systematically enforcing control on US based businesses in a way that is historically more synonymous with, say, China.
A lot of us worry it is the former. I wonder if it's better or worse if it is the latter. The US, having seen that Europe and the world seem to tolerate interference by CPC based on opinions expressed by Alibaba or Chinese car companies, may have decided that it's fair game.
digitaltrees 22 hours ago [-]
Actually they’re predictable: loyalty and submission to trump personally. Every thing flow from that. It’s not the rule of law.
zaptheimpaler 21 hours ago [-]
So far the other countries are doing a lot of talking and big government initiatives but nothing has really changed. All the best talent still moves to the US, all the VCs are still unimaginative losers who want 5 meetings to give you $50K, the industry titans are still sclerotic and unable to distinguish their head from their ass, the culture doesn't support risk-taking or innovation. The biggest boom of the next 10 years is AI and it is firmly HQ'ed and centred in the US, as usual. The EU had a predecessor to the Chips Act in 2013 [1] that went nowhere, then a Chips Act that went nowhere, and now a Chips Act 2.0! It would be great to have more options but it doesn't seem like other regions outside China are capable of doing anything at all.
I think this can be said in many contexts these days.
davidw 1 days ago [-]
They don't care about the long term - it's a smash and grab to get as much as they can while they can.
advisedwang 21 hours ago [-]
> Is regulation unregulated in the US?
For the most part no. The Administrative Procedures Act outlines a fairly robust process for rulemaking (ie regulations must go through a notice period, must be justified etc) and adjudication (ie per-person decisions must have notice, ability to have counsel, be appealable etc).
However there's also lots of authority given to the president (and executive office of the president) to which the APA doesn't apply. This includes powers the constitution gives to the president directly and (mostly emergency) powers congress gives to the president. Sadly political processes and supreme court cases have hugely increases what can be done with this*.
Moreover, it's been a frequent pattern (over many decades, but especially under Trump) for the president to simply nakedly overstep their authority. Sometimes there's no push back (see: 9/11); sometimes the damage is done even if it's eventually overruled. Never is there any consequences though, so the pattern repeats.
* E.g. The president's control of the military is supposed to be checked by congress deciding when to declare war. However the war powers act has in practice played out to allow the president to unilaterally start wars, especially as the supreme court has stepped in to make the war powers act less effective.
andai 1 days ago [-]
Never interrupt your opponent while he is in the middle of making a mistake. –Sun Tzu
zombot 13 hours ago [-]
> I don’t see us as the US on the right path.
Is anyone still assuming the administration's actions have anything to do with US national interests?
ryandrake 21 hours ago [-]
Unfortunately, in the USA, we are stuck with two parties. Measured along axes of competence and capability of taking action: The (D) party is competent+incapable, and the (R) party is incompetent+capable. So every four years we get to choose between governance that knows what it is doing but too paralyzed to act, and governance that doesn't know what it is doing but acts with impunity.
guelo 21 hours ago [-]
I don't think that's true about Ds being incapable. For example Biden's Inflation Reduction Act has been very successful at boosting clean energy industry, creating jobs, lowering emissions. Including huge benefits to red states. The problem is a lot of times Democratic solutions, like Obamacare, are too technocratic, or the benefit is too diffuse for people to notice. There's also the issue of the huge growth of the right wing's propaganda power successfully hiding any D achievements from voters.
quantified 21 hours ago [-]
D's absolutely suck at messaging, and they are not that good about cohesion. Of course, being overly cohesive leads to problems as we see in today's world.
amazingamazing 1 days ago [-]
> We’re proposing stronger regulation of the technology. We’re proposing giving the government the ability to, again in a narrow way, block deployment of unsafe technology.
Okay, so why are other models not banned too? This "jailbreak" works for them as well.
soerxpso 17 hours ago [-]
Anthropic has alleged that this model is much more dangerous than other currently available models. Their CEO has said so publicly multiple times. It's like asking why cesium isn't banned if nuclear missiles are banned.
(whether Mythos is actually that dangerous is beside the point; considering that Anthropic claims that it is, it makes sense to regulate it)
tiahura 22 hours ago [-]
How would you know? Amazon and the NSA thought it was a problem, but you know better?
enraged_camel 19 hours ago [-]
Know what better? They told us what the "jailbreak" entails.
Lionga 1 days ago [-]
Anthropic CEO: REGULATE AI (so we don't have competition).
Goverment: ...
Anthropic CEO: NO, NO, NO NOT LIKE THIS.
sanderjd 1 days ago [-]
Yes, correct. Singling out companies based on political beefs and to incentivize everyone to curry favor is the worst way to do regulation.
matheusmoreira 23 hours ago [-]
You're correct, yet it's also a fact that Anthropic was attempting regulatory capture in order to limit open weight models, cripple their competition and solidify their market position. Nothing wrong with enjoying the sheer Schadenfreude of the situation. Their self-serving fearmongering had the most hilariously unexpected result possible.
sanderjd 22 hours ago [-]
Maybe. But I disagree in general with "nothing wrong with enjoying the sheer schadenfreude of the situation". Schadenfreude is a description of a common human impulse that is not a good impulse!
matheusmoreira 21 hours ago [-]
Normally I would agree with you, but in this case I'm gonna make an exception. These are billionaires who campaign against open weight models. Their misfortunes are our freedom.
sanderjd 13 hours ago [-]
Nope, if everyone makes exceptions for the people that specifically annoy them, then there will always be some exception for everybody. Much better to endeavor not to give into your own vices, even the understandable ones.
You can come up with and advocate for a policy to be evenly applied to all the billionaires whose "misfortunes are our freedom", instead of supporting this corrupt favoritism.
Lionga 1 days ago [-]
Anthropic CEO is full of shit, got what he deserved.
He told the world about 20 months ago that all software developers would be gone in 12 months. Why has he not fired all Anthropic Devs yet?
sanderjd 1 days ago [-]
You can be annoyed by a person or company without wanting the federal government of the United States to grind a personal axe against that person or company.
What you're advocating for here is banana republic governance. It's bad.
digitaltrees 22 hours ago [-]
Remember arbitrary application of law expands and affects everyone. You will be affected. The advantage of the rule of law is the equal application to everyone.
tiahura 22 hours ago [-]
Isnt the situation sui generis?
sanderjd 21 hours ago [-]
No, it is not. This is the entire thing about the rule of law. People constantly think their pet thing is sui generis. But actually the even handed application of the rule of law is a general good.
tiahura 21 hours ago [-]
Argue with Anthropic, Amazon, and the NSA. They all say it is.
sanderjd 13 hours ago [-]
I disagree with them, but they aren't the ones I'm discussing this with here. I'm discussing it with you. What do you think?
tiahura 3 hours ago [-]
I think none of us know the real facts.
digitaltrees 19 hours ago [-]
The burden is not on them to show that. It’s on the government to show why the restriction is necessary to achieve a compelling public interest.
I’ll make them argue that when you get Sam Altman to admit he lost the lead to anthropic
tiahura 3 hours ago [-]
It’s on the government to show why the restriction is necessary to achieve a compelling public interest.
Perhaps, ITAR and the like isn't my legal specialty.
Isn't on a company's CEO to make sure everything is copacetic with its public and private partners?
digitaltrees 20 hours ago [-]
The government has an affirmative obligation to show why it is unique then.
InsideOutSanta 24 hours ago [-]
> Anthropic CEO is full of shit, got what he deserved.
This is an incredibly dumb way to judge a government's behavior.
b0sk 1 days ago [-]
During the Obama administration, when solar companies were subsidized, the refrain was that "government was picking winners and losers". But now, there's a cut and dry case of one company getting targeted by the administration. Either ban all the powerful models or ban none of them.
digitaltrees 22 hours ago [-]
Regulation is a published rule applied equally to all the industry. Not one applied to a target for retribution or to extract submission
virissimo 22 hours ago [-]
Applying equally to all the industry is a characteristic of good regulation, not of regulation tout court.
digitaltrees 20 hours ago [-]
Arbitrary enforcement of a regulations is against the principles of administrative law and gives rise to a cause of action that can set aside the governments action. So all regulation has that as an explicit assumption.
There is also a case that it violates due process because anthropic wasn’t provided adequate notice before enforcement or an opportunity to challenge, respond, or take steps to comply.
There is latitude for compelling state interests and national security is one with the widest latitude and potentially one that counts can’t review. But the court of public opinion may draw the inference that this is direct retaliation for refusing to make the models available for autonomous weapons and spying.
wrs 22 hours ago [-]
Anthropic published, not a two-word request for regulation, but a lengthy proposal for how regulation could work, with safeguards and due process.
Instead, one morning the administration just woke up, grabbed the nearest hammer, and went smash, like it usually does.
christkv 22 hours ago [-]
They proposed a plan to ensure their dominance of the space and to cripple any competitors using regulation as the weapon. So I’m not going to feel sorry for them and their temporary plight
mmooss 21 hours ago [-]
Do you have evidence of that?
22 hours ago [-]
jasonlotito 1 days ago [-]
PAGE NOT FOUND The page you requested was not found. You may have followed an old link or typed the address incorrectly.
I'm sure it says something that you think makes this a slam-dunk rebuttal, but you should fix your link first.
Edit: Saw the link. I was right. Nothing in that link makes a difference.
sailfast 1 days ago [-]
Any company building this kind of tech should have a separate shell headquarters on some low regulation island country.
Any government work can be done via MOU with a US subsidiary staffed by Americans.
This kind of capricious, unexplained control is bullsh*t.
I’m not saying that I want companies to have to go offshore or that that would be a good thing. Just that you’ve got no leverage if your corporate structure can be destroyed on a whim. This goes for any company reliant on a whimsical executive branch.
While there could still be fights over the technology and the company, a tech provider would still be able to serve other customers and have more leverage.
tomjakubowski 1 days ago [-]
A separate shell headquarters on some low regulation island country? Like FTX?
blitzar 21 hours ago [-]
They should also have a company penthouse / polycule in said low regulation island country.
And drugs, lots of drugs.
tiahura 22 hours ago [-]
Which would guarantee 0 VC investment once the itar notice goes out.
rootusrootus 1 days ago [-]
> Any company building this kind of tech should have a separate shell headquarters on some low regulation island country.
Like where? How about Cuba? Where Guantanamo is, where we put terrorists because we don't want them to be subject to US laws (or more specifically, we don't want ourselves to be subject to those laws when we are dealing with said terrorists). How about no.
International law has only as much power as the biggest military willing to enforce it. Hiding on some little island is not at all a good strategy for trying to evade the US government.
snvzz 12 hours ago [-]
>Any company building this kind of tech should have a separate shell headquarters on some low regulation island country
That would seem to me as grounds for piercing the corporate veil and holding the people behind it directly responsible.
High treason or some similar charge. Capital punishment not out of consideration.
pitched 24 hours ago [-]
Without going that far, send a few people a couple hours North instead and serve international customers from Canadian data centers. As far as I understand, it is only blocked in the US, right?
breakpointalpha 24 hours ago [-]
It's routine and trivial for the US Gov to enforce export controls on non-US based companies.
This "across the border" strategy will not work.
Seagate Singapore was hit with civil penalties and settlements for shipments to Huawei under the Foreign Direct Product Rule.
ZTE, a Chinese company, was criminally prosecuted and settled with U.S. DOJ/BIS in 2017 for violating U.S. export controls and sanctions.
davidfekke 21 hours ago [-]
None of the best minds in AI Research are from the United States. We would not be were we are without the work of foreign scientists. There are more STEM students in China than we have students in the United States, and those universities don't give out automatic As like we do here in the United States.
sailfast 1 days ago [-]
Misleading headline based on one off-hand slack comment and quotations from experts outside the firm.
ajju 1 days ago [-]
Unfortunately saying it in the press is going to make it even harder for them in the current environment. I know this administration gets non-trivial support from the valley, but what to most outsiders seems like "targeting businesses based on personal vibes" is going to do long-term harm to the U.S.
I hope those who represent technology in government, especially the AI head David Sacks are giving this due consideration.
khalic 1 days ago [-]
Ah the good old “stop resisting” falacy
ajju 23 hours ago [-]
If you re-read my comment, I am not saying that.
It seems to be observably true that expressing opinions critical of the administration, even about one's own business, leads to further harm from government right now in the U.S. in a country which is supposed to have strong freedom of speech rights.
It's obviously unfortunate that this is the case.
It's also a shame that this is happening when the folks regulating AI have a tech industry background.
The administration seems to be clearly signaling "stop resisting". I am not sure how you read me as saying that.
martythemaniak 20 hours ago [-]
I think you are misreading this situation. The Musk-Thiel cluster, of which David Sacks is a major member, is one of the main groups trying to destroy the rule of law and turn the US into a personalist regime. They are doing this because they think they'll be major beneficiaries of the new regime. As they see it, the long term success of the US, ie the rule of law, is against their personal interests.
seviu 1 days ago [-]
Anthropic employees are right, but maybe this is for good. It certainly has opened my eyes.
I can’t rely on using a technology that the US administration can ban at will.
IMO without getting into personal thoughts about how capable the current US administration is, last Friday move sent a very powerful signal to the industry.
Also I don’t think China releasing so many good models, capable to compete with Opus 4.8 and GPT 5.5, all at once, is a coincidence.
nickff 1 days ago [-]
Are you saying that you think the US government is unpredictable and arbitrary, but that the People’s Republic of China is not? Do you remember all the PRoC’s strange and sudden policy shifts (e.g. steel, real estate, education, football/soccer, etc.)?
It seems to me that in the case of AI (as with many other modern technologies), you rely on vendor/creator support and updates to stay relevant, so the ‘next’ model matters more than the current one, and we have no idea whose next model will be open (and whose won’t).
digitaltrees 1 days ago [-]
I interpret the comment as saying there are lots of viable models and it's now crystal clear personal cloud or local Ai is the only reliable path.
baq 1 days ago [-]
Not OP and I wholly agree, but you can’t dismiss the fact that they are releasing those weights. Their agenda is quite obviously to make Anthropic and OpenAI CFOs sweat bullets, but it isn’t our problem as AI consumers, right?
nickff 1 days ago [-]
Yes, I agree that it is possible that the 'open source model providers' are doing the equivalent of 'dumping' in an attempt to establish a dominant market position, or at least a foot-hold. I am generally a skeptic when it comes to the effectiveness of 'dumping' as a long-term strategy (as the producer tends to hemorrhage consumers when it increases prices), but some may see it as problematic.
fnordpiglet 1 days ago [-]
Open weight models don’t allow central oversight. That’s the difference.
nickff 1 days ago [-]
If you’re happy to use the current one forever, then yes. I was amending my comment above to address this when you posted yours.
fnordpiglet 16 hours ago [-]
I think for many practical purposes the frontier open weight models are almost universally good enough for most things. There may be greater and greater frontiers but at q certain point it becomes like IQ. Having a 150 IQ doesn’t mean you’ll be more successful at any particular task over someone with a 125 IQ. Indeed there’s a diminishing return on intelligence on many utility functions where being more intelligent yields more be same or worse ultimate outcomes. It might very well be the person with a 150 IQ could understand some extraordinarily complex and esoteric concepts faster, but it doesn’t mean with more effort the 125 IQ person can’t either; and sometimes that extra time spent yields better outcomes overall.
I suspect AI will be somewhat similar where even if the linear scaling laws continue to hold the practical utility of a model flattens for almost all conceivable use cases.
In some ways I already feel this has begun to happen. The marginal utility of opus class models and fable has in my perception begun to flatten. While I can tell the differences they aren’t earth shattering. I could continue to use the present models for the rest of my life and be ludicrously more productive simply by adapting within their constraints through ever more sophisticated applications.
What holds back the open weights IMO is hardware scaling and industrial production. As the enormous transfer of wealth in debt and equity markets unfolds with semiconductor and adjacent companies and the corresponding capital investments are made, and the eventual bubble pop leading to over capacity and market flooding, as well as advances in technology, math, techniques, and efficiencies, will make very large open weight models more directly attainable. This will also lead to chimera models that MOE very large models to get very close to the 1-2T parameter dense models, at which point I suspect utility for almost all uses is nearly fully saturated.
There will be areas where more capable models are needed but they will be frontier models on frontier problems. This, IMO, is inevitable, and without some criminalization of weights (see the attempts to criminalize encryption algorithms in the 20th century and all the wonderful tshirts that emerged). It’ll be harder to print a trillion parameter model on a shirt but I’m sure someone will try, as will governments try to keep us in our boxes slaving for food coupons and basic rights like health care.
thatmf 1 days ago [-]
> Are you saying that you think the US government is unpredictable and arbitrary, but that the People’s Republic of China is not?
Why not both?
That seems the crux of the state we're currently in; what daylight there was between the two is quickly fading.
spelk 1 days ago [-]
>Do you remember all the PRoC’s strange and sudden policy shifts (e.g. steel, real estate, education, football/soccer, etc.)?
I didn't realize I could download a Shanghai apartment.
seviu 1 days ago [-]
Right now I rely on whoever that is opensourcing the models.
I wholeheartedly agree with what you said about China.
But I can’t shrug off the fact that fable was taken down within minutes for reasons that are childish and petty.
I am sorry but I can’t use any US AI if I don’t have the guarantee that I will be able to use it tomorrow.
And Trump showed us he is willing to take it out whenever he wants.
An opensource model on the contrary, I can host myself, or use a miriad of providers, mostly non chinese.
jmaw 21 hours ago [-]
> I am sorry but I can’t use any US AI if I don’t have the guarantee that I will be able to use it tomorrow.
To be fair this is every commercial model. We have already seen GHCP increase prices by anywhere from 10-100x (depending on usage). And old models get retired all the time. While these are not exactly the same as a cutting edge model being shut down, increasing prices a super high amount leads to effectively the same outcome.
UncleOxidant 1 days ago [-]
> And Trump showed us he is willing to take it out whenever he wants.
Yes, the actions of this administration on Friday should have sent shockwaves through the market - a market that's currently "high on AI". How do you get a return on all of that AI investment if the administration can jump in at any time and say "Nope, you can't use this very advanced model!"? (the Iran "deal" over the weekend, I think helped cushion that blow, but eventually it's going to sink in)
khalic 1 days ago [-]
Yep, any American closed model is now a de facto existential risk for any company relying on them.
The latest open models are so good it’s worth the 6-8 months delayed capabilities. At least for coding
fnordpiglet 1 days ago [-]
The problem is there’s a real wall on the vram side. While fused main memory is ok the inference speeds on larger models are impractical. With vram on a GPU the machine class, power requirement, GPU costs, and other factors put them out of most people’s reach. Cloud GPUs require a second job to keep available and hot. What closed providers offer is packing and scale advantages as well as infrastructure. The scaling laws here aren’t the same as Moore’s law - in fact they predict more required hardware and more scale over time. Moore’s laws isn’t keeping up with expanded needs and the ability to fab and produce at scale the specific things that weren’t needed a few years ago are lagging. So it’s not a 6-8 month lag; it’s a lag that will be induced by hardware scarcity and an ever increasing lag until something fundamentally changes with matmul.
wahnfrieden 1 days ago [-]
I will use the best available while it is available. 8 months ago with Codex would be intolerable today.
khalic 1 days ago [-]
On a personal level, ditto. But at a business level, this kind of uncertainty will kill you. You need to be able to plan ahead.
realusername 1 days ago [-]
I believe we have somewhat plateaued and each percentage gained seems to be an exponential effort.
Fable was around 10x GPT5 pricing and 100x Chinese models pricing, was it really 100x better? I Don't think so.
If you want a personal story, I just solved a complicated coding problem with Kimi 2.7 that GPT 5.4 failed with.
wahnfrieden 9 hours ago [-]
5.5 is far ahead 5.4. I don't see any plateau (from using these 16+ hours a day 7 days a week)
enraged_camel 1 days ago [-]
>> I can’t rely on using a technology that the US administration can ban at will.
And you think China will not do the same thing if their models ever become genuinely frontier-level?
khalic 1 days ago [-]
Then the US will publish their own open weights to outmanoeuvre china.
What’s intolerable is having a tool that’s subject to this risk.
So open models it is
1 days ago [-]
mpalmer 1 days ago [-]
Good luck with your trade secrets I guess
khalic 1 days ago [-]
You can run the Chinese models on your infra, most are open weights. Not saying it’s out of the goodness of their heart, but the fact is, they’re open.
quantified 21 hours ago [-]
It could be as simple as tipping the scales towards Grok.
ghusto 23 hours ago [-]
The Cloud Act instigated the EU's migration away from companies in the USA, and things like this lit a fire under our butts. You should see how companies here are scrambling to move their data and compute to the EU.
The impact will last long, long, after this administration is gone. Trump will have his legacy.
spelk 1 days ago [-]
One wonders if this might be a net positive for the world if Anthropic is forced to terminate it's non-US citizen employees.
I would hope they return to their home countries with their expertise and start or join new competing frontier labs, similar to how Taiwan's homegrown semiconductor industry arose from US companies enforcing a bamboo ceiling on their Asian engineers. Taiwan was able to repatriate their nationals and make incredible compounding leaps in the semiconductor industry, to the loss and chagrin of the US. [1]
I can't imagine how you'd restrict access to these models to USA nationals only. Every company in the USA would have to verify the id of every employee's model use. What do you do about programmatic AI workflows? How does Anthropic even continue to function internally right now on this Mythos and the future model development?
ifwinterco 1 days ago [-]
Indeed, forcing Europeans to stay in Europe might be the only thing that could revive the moribund European AI scene
khalic 1 days ago [-]
The brain drain goes in the other direction champ. Scientists and engineers are already leaving the US en masse because of the current situation
inglor_cz 22 hours ago [-]
Someone like Saudi Arabia may attract them with untaxed high salaries instead.
bflesch 1 days ago [-]
Nothing of value is lost for Europe because we mostly work with natural intelligence. The malinvestment bubble happens in the US, and hardware prices will come down.
We're all dependent on Taiwan but others also depend on ASML which is located in Europe. Available AI services are good enough, and we sit and watch while US leadership and population is beta testing AI mass surveillance for us.
In terms of AI-based warfare Ukraine is very much ahead and they are also European for which we are very thankful.
So feel free to keep your AI scene and AI hustlers in the US, we don't need them.
ifwinterco 10 hours ago [-]
I am British and have no love for the US right now but in terms of talent it’s notable how many prominent researchers at the top US labs are originally European, and the contrast with the top European labs (of which there are ~none).
Does this actually matter as much as some people think? Probably not because LLMs won’t lead to AGI, but it is indicative of a broader problem
bflesch 10 hours ago [-]
US startups are trying to do a commercialization of AI, but their marketing is about being a "lab" and doing "research" - which they are not. They're trying to commercialize the hell out of the LLM approach, while fundamental research is still happening at universities.
ifwinterco 9 hours ago [-]
Fair point, but the issue I’m getting at is all too often from a European perspective the pipeline is:
Expensive publicly funded European education for 25 years -> high paid job in the US.
Costs are all paid by Europe, tax revenue and job creation all accrue to the US.
If we’re not careful it ends up being a straight transfer of value from ordinary European taxpayers to the US government which seems… suboptimal.
If the US can short circuit that process by shooting themselves in the foot, Europe will benefit
bflesch 9 hours ago [-]
Fully agree but it is still a risk for the US, because if the clever immigrants notice they can't get as rich as their American-born peers and on top of that US leadership publicly antagonizes their country of origin, those foreign-born engineers can quickly become foreign agents.
ifwinterco 8 hours ago [-]
Yes true, and the US is maybe big enough to adopt that approach.
It’s what China does (except for poaching people from Taiwan), but they have 1bn people so for them it makes sense
inglor_cz 22 hours ago [-]
"In terms of AI-based warfare Ukraine is very much ahead and they are also European for which we are very thankful."
Note that the Hornet drones which are now devastating Russian logistics behind the lines are US manufactured.
bflesch 9 hours ago [-]
Source? Even if this is true it is not noteworthy, because the damage done by US to their allies far outweighs such small "help".
From a peaceful office in Germany, it might not be. For a country fighting for survival, the perspective is different. I don't think you could persuade them to give up Hornets right now, in the name of abstractly getting back at the US in general. (With whom they have a lot of bad blood!)
bravetraveler 1 days ago [-]
I'm accusing Anthropic of being facetious
FloatArtifact 1 days ago [-]
Frankly, they've been asking for the AI industry to be regulated. They just didn't expect it would be just them. Although I'm going to say here now that I think this is going to have a broader ripple effect all the way down to open source and foreign countries developed models.
12 hours ago [-]
blitzar 1 days ago [-]
> Workers at the artificial intelligence company have been puzzled and increasingly concerned by the administration’s move to limit their latest A.I. models.
There is no way any employees at anthropic are this dumb.
sanderjd 1 days ago [-]
"puzzled and increasingly concerned" is just journalist speak for "super pissed off".
tmaly 23 hours ago [-]
Philip Zimmermann
semiinfinitely 21 hours ago [-]
sounds like they got what they asked for
nailer 22 hours ago [-]
Amazon AI and the government is right in saying Fable infosec protection is easily crackable.
I used Fable last week to scan my own work for security errors (and readability issues), I wasn't thinking about the ban on infosec use I just wanted the most well trained output.
One project in the monorepo triggered a Opus fallback and infosec/biology warning.
The other project scanned fine and showed a bunch of potential exploits.
vasco 22 hours ago [-]
[flagged]
vasco 12 hours ago [-]
Lmao and this gets flagged. No shame.
samlinnfer 1 days ago [-]
Anthropic is just the whipping boy. It keeps everyone else in line.
extr 1 days ago [-]
I mean obviously they're correct but also the complaints of the administration aren't totally without basis.
- They're obviously being targeted politically because they refuse to kiss the ring, vibes, whatever you want to call it.
- They're also justifiably being scrutinized because they just spent like 3 months telling everyone that Mythos is a nuclear bomb and telling the government to fuck off as they drip fed access to a bunch of random corporations.
sanderjd 1 days ago [-]
Yes. It's true that the pretext is not entirely without merit. But it's also true that it's a pretext.
But in cases like this, the pretext shouldn't be taken too seriously. Because they would have just found another one. It isn't actually particularly important that it is plausible, it's more like a happy coincidence.
Ataraxic 1 days ago [-]
The amount of self owning that is happening to AI companies is crazy.
I don't understand why their marketing department/execs can't see the conflict between claiming AI is going to take all jobs, that the model is super dangerous and AI hate among the general public, increased governmental oversight.
Anthropic is guilty of the latter but the former applies to most of these AI companies.
ekidd 1 days ago [-]
As far as I can tell, much of Anthropic genuinely believes that someone will build an AI in the next 3-20 years that's significantly smarter than any human alive. Sounds wild, but a lot of their people have been saying this since 2018 or even earlier. I think they're true believers. Furthermore, they believe that building such an AI would be dangerous.
So their plan is:
1. We can't stop other people from building something dangerous.
2. But we can get there first.
3. If we build it, it has maybe a 15% chance of killing everyone alive. (I think that's a number I've seen Dario use before, but I may be wrong.) If OpenAI or China build it, the odds would be worse.
Obviously, if Anthropic is actually correct about (1) and (3), then nobody should allowed to build frontier AI.
People find it really hard to believe that (a) anyone believes in the possibility of dangerous AI in our lifetimes, and (b) that someone could believe what Anthropic seems to believe and then still go ahead and gamble with everyone's lives anyway.
extr 1 days ago [-]
To me they are genuinely trying to walk a tough line - they legitimately believe that they need to warn the public and make a lot of noise so society can try to adapt to this technology. OTOH no adaption (good or bad) can take place if the models themselves are so restricted as to be inaccessible, or if the powers that be don't understand it well enough to put the right policy/laws in place.
22 hours ago [-]
khalic 1 days ago [-]
Ah yes, giving cybersecurity a 3 month advance is such a horrible thing to do!! Bad anthropic
extr 1 days ago [-]
I think they were totally correct in spirit. But RE: details about them giving access to an SK corp with possible Chinese ties. Of course that raises eyebrows in the USG, justifiably. Sloppy work from Anthropic.
khalic 1 days ago [-]
No this is a pretext, nobody who is following the matter of their feud with Hegseth. Trump tweeted they’ll regret it. Textbook retaliation
This has some insight in my many cyber security professionals want Fable/Mythos open.
1 days ago [-]
martythemaniak 1 days ago [-]
Yes, obviously. People really need to update their priors on how the US operates now. Laws are out, loyalty is in. There is an extraordinarily powerful unitary executive in whom the will of the people resides and it is the job of the government and society to work towards the will of this powerful executive. There are no checks or balances or alternative centres of power (civil, political, clerical, etc) allowed.
This particular executive loves money, praise, and submission. When Anthropic submits (eg agreeing to whatever the DoD demands), issues public praise and makes appropriate donations (eg ballroom, memecoin, etc), then they can do as they please.
Students of history will find this new MO very familiar and very depressing.
Johnny_Bonk 1 days ago [-]
The most depressing part is how few people are students of basic civics, history and political science. We truly live amongst house cats that don’t know how good they have it and have begun dismantling the very house that they’ve lived properously in for so long. (Obviously not including historically and currently marginalized etc)
sanderjd 1 days ago [-]
People don't need to update their priors to accept that this is the new way, they need to recognize the reality of the current situation and also its aberrance.
nailer 21 hours ago [-]
> Laws are out, loyalty is in.
Anthropic themselves refused to allow the Pentagon use of its models for any legal purpose - instead the wanted a subset picked by Anthropic. It's ironic that you're accusing the government of ignoring the law when Anthropic openly wishes to set it's own private laws.
etchalon 1 days ago [-]
[flagged]
garciasn 1 days ago [-]
Petty means they're excessively concerned with trivial matters, minor slights, or small grievances; the Trump Administration is FAR MORE than petty.
They're not simply petty; instead, they're vindictive, retaliatory, and vengeful.
sanderjd 1 days ago [-]
Porque no los dos.
jimmydoe 1 days ago [-]
[flagged]
sanderjd 1 days ago [-]
Genuinely, if the policy is that this capability is too dangerous to be in private hands, and thus all "frontier" labs must be publicly owned, that would be a defensible policy (that I would disagree with). But picking and choosing preferred companies like this is just indefensible gangster stuff.
teling 1 days ago [-]
That would be great if that could lower API costs
petilon 1 days ago [-]
[flagged]
seydor 1 days ago [-]
Trump vs anthropic
Elon vs openai
Trump vs elon
Bezos vs anthropic
When movie?
thewebguyd 1 days ago [-]
Still waiting on the Elon vs. Zuck fight.
matheusmoreira 23 hours ago [-]
Movies aren't as entertaining as reality!
neko_ranger 1 days ago [-]
I'm waiting for the AI: Endgame
esafak 1 days ago [-]
What do you need a movie for; we're living it?
sanderjd 1 days ago [-]
The movie version will hopefully be more fun. My experience of the rise of Facebook was dull and depressing. But The Social Network is a great film.
pqdbr 1 days ago [-]
So all non-fictional movies ever made were unnecessary?
andxor 1 days ago [-]
Another opinion piece?
bannable 1 days ago [-]
This is not an opinion piece, which is clear as soon as you open the article.
andxor 1 days ago [-]
I'm sorry.
viccis 1 days ago [-]
Anthropic sowing: Hehe yeah guys check out this model it's super duper scary and powerful it's a huge deal
Anthropic reaping: Hey wait a second you weren't supposed to take that seriously it was just marketing :(
1 days ago [-]
reasonableklout 1 days ago [-]
It can both be true that models at Fable capability level are national security concerns while Anthropic is being hypocritical.
If the Trump admin is also willing to apply the same scrutiny to GPT-5.6 and other Fable-level models, I think it is a good thing. But given the admin's history with Anthropic (such as declaring it a supply chain risk while ignoring Chinese labs), there is some smell of targeting.
viccis 19 hours ago [-]
I'm not saying they're hypocritical. I'm saying that this would be far less likely to have happened had they not advertised its power by bringing up its danger.
ReptileMan 21 hours ago [-]
Except they can't - mostly because Sam Altman is savvy enough to not shout from the rooftops the national security implications. The government can't regulate models because models are unmeasurable right now. And whatever test USG can create can be benchmaxxed in reverse eventually.
So at some point - don't say your model is capable of hacking even if it is and you are clean.
godwinson__4-8 1 days ago [-]
Was it not just a few weeks ago the government was telling Anthropic to take the guardrails off of the "super duper scary" models?
Now they want them back on?
AI for me and not for thee?
Or just another opportunity for a petulant administration to target Anthropic for not doing what they wanted the first time?
8note 1 days ago [-]
literally yes. that is what export restrictions do
viccis 19 hours ago [-]
>AI for me and not for thee
100%, I mean we already have Mythos vs Fable
tiahura 1 days ago [-]
If you manage the corner coffeeshop, and the city health director calls with an urgent matter, you take the call and assure them you'll take care of it.
Calgaryp 1 days ago [-]
They just removed Fable in France
catigula 1 days ago [-]
They did a whole PR cycle about how dangerous their technology was. They refused to release it for public consumption.
How is this “targeting”? It’s literally what was requested.
InsideOutSanta 1 days ago [-]
They were asking for regulation. They were not asking to be singled out and prevented from releasing newer models while everybody else keeps going as usual.
Based on the reasoning for blocking Fable, every current model should be blocked. GPT 5.5 is similarly strong and has fewer guardrails, for example.
crispyambulance 1 days ago [-]
> They were asking for regulation.
That's what they said, but were they, really?
I've heard others call it a marketing ploy that AI companies keep coming back to-- the idea that "it's soooo dangerous, we have to be careful".
thewebguyd 24 hours ago [-]
> That's what they said, but were they, really?
Yes, most likely, but not in this form, obviously. They want open weight models regulated for regulatory capture, and I'd assume they want an actually documented framework applied equally across all labs.
If GPT5.5 has the same capabilities Fable did, then for consistency sake, it should have also been subject to this ban.
Regulate or not regulate, but the government should not pick winners and losers.
crispyambulance 20 hours ago [-]
> Regulate or not regulate, but the government should not pick winners and losers.
Indeed, it has been very clear since Jan 20 2025 that our executive branch of government is not a rational actor. I don't expect them to make the right decision about anything.
InsideOutSanta 1 days ago [-]
I can't see into their heads, and I also don't think it matters whether they were making a good-faith argument or lying through their teeth. The fact remains that what is currently happening is not what Anthropic asked for.
ReptileMan 21 hours ago [-]
When the Gods wish to punish us, they answer our prayers
ReptileMan 21 hours ago [-]
I don't know. I have heard from reliable source - aka the CEO of antrhopic that Mythos is in league of its own. And that the rest are grossly inferior. Anthropic can't have it both ways. If they are special - only them should fall under regulations. If they are not - why have you committed fraud in the last couple of months right before your IPO saying otherwise.
matheusmoreira 1 days ago [-]
That's hilarious. They aren't even wrong, but they asked for this. They accuse Trump of targeting them? I accuse Anthropic of targeting open source models with outright banishment in order to solidify its own economic position. They tried some "enlightened" variation of regulatory capture, but didn't dance to Trump's tune, so Trump gave them the regulation they wanted so much.
spullara 1 days ago [-]
Anthropic declares: "Mythos is too dangerous too release to the public"
Proceeds to release Mythos plus safety guardrails as Fable.
Amazon removes guardrails from Fable, getting access to Mythos.
Government takes Anthropic's word for it and tells them to pull it until the guardrails can't be removed. They refuse. Government forces them.
sanderjd 1 days ago [-]
> Amazon removes guardrails from Fable, getting access to Mythos.
This doesn't seem like an accurate description to me. I think something like "Amazon demonstrates a jailbreak of one class of Fable guardrails" would be a more accurate description.
It doesn't even really mess up your narrative to state it accurately, but your choice of a more hyperbolic statement brings into question the good faith of the narrative you're painting.
tiahura 1 days ago [-]
We really don't know for sure exactly what Amazon did. They're being quiet, the other parties aren't trustworthy, and the reporters mostly don't know what they're talking about.
I would argue that that article describes a demonstrated bypass of one class of Fable's guardrails.
khalic 1 days ago [-]
Yeah if you ignore the fact the Us government retaliating about Anthropic not wanting their AIs in weapons systems.
rootusrootus 1 days ago [-]
Was Fable really the full Mythos model but with guardrails added? I had assumed Fable had a reduced parameter count or something, like a Sonnet to an Opus. Interesting!
jasonlotito 1 days ago [-]
No, the parent is lying. Anyone suggesting that needs to bring full receipts, but they can't. They are liars at best.
InsideOutSanta 1 days ago [-]
Amazon did not remove any guardrails from Fable.
0xy 1 days ago [-]
What? I personally experienced Fable outright refuse to do ANY security-related tasks, including hardening code or modifying security-related features. That was a guardrail. It was bypassed.
Anthropic themselves specifically called them safeguards. [1]
"When Fable’s classifiers detect a request related to cybersecurity, biology and chemistry, or distillation, the response is automatically handled by Claude Opus 4.8 instead"
This is exactly what was bypassed. They got Fable to work on security topics.
They got Fable to fix bugs, including security issues, which is what it is supposed to do.
0xy 24 hours ago [-]
What? Fable was designed to refuse to work on security issues, as Anthropic specifically confirmed. How is forcing Fable to work on things behind guardrails not breaking a guardrail?
This is Anthropic's own claim. They were very specific. Have you read their own claims?
InsideOutSanta 24 hours ago [-]
Yes, I have read their own claims. Here's the relevant part:
"When Fable’s classifiers detect a request related to cybersecurity, biology and chemistry, or distillation, the response is automatically handled by Claude Opus 4.8 instead. Users will be informed whenever this occurs."
Asking Fable to fix bugs in a code base is not "a request related to cybersecurity." When Fable was asked to fix bugs and then proceeded to fix bugs, that was not "removing guardrails". Fable did exactly what it should have done. Claiming otherwise makes absolutely no sense at all.
0xy 23 hours ago [-]
Fable specifically refused to harden the security of codebases. If you use misdirection to force Fable to do just that, that's the removal of a guardrail.
Anthropic specifically stated that ANY security requests should be shunted to Opus 4.8. This was bypassable.
I don't see what your confusion here is. Fable was prevented from working on any security tasks. A significant amount of people, myself included, witnessed Fable refusing to harden code as a result. Bypassing that is a bypass of guardrails.
Your assertion that working on security is not working on security because you used misdirection is of course, preposterous.
You wouldn't be making the same claim if Fable refused to work on chemical weapons research but happily proceeded to do so if you claimed it was for eradicating pests.
InsideOutSanta 9 hours ago [-]
> If you use misdirection to force Fable to do just that, that's the removal of a guardrail.
Asking a model to fix bugs is neither misdirection nor a security request.
> I don't see what your confusion here is.
That's because I'm not confused :-)
> Fable was prevented from working on any security tasks
I don't think that's true based on what Anthropic said, and I also don't think it can be true.
What do you propose Fable's behavior should be if you ask it to fix bugs, and it encounters a security issue? I'm assuming your solution is that when you ask Fable to "fix bugs," and it encounters a bug that could be exploited as a security vulnerability, it should fall back to 4.8. But that doesn't solve the problem, because as a user, I can now see where that occurred, so I still know where the vulnerability is. That's not substantially different from the current outcome, where it just fixes the bug.
It would also mean that Fable could barely make it through any code review without falling back to 4.8, because almost any non-trivial code base has aspects that could be interpreted as security vulnerabilities.
The alternative would be for the model to use its hidden thinking to decide not to fix the bug, but that seems even worse.
1 days ago [-]
giancarlostoro 1 days ago [-]
This. They also wanted more regulation around AI. I'm guessing they're no longer quite as interested in this.
InsideOutSanta 1 days ago [-]
Blocking one model is not regulation. The comments on this so far are wild.
sanderjd 1 days ago [-]
... it isn't? What word would you use for a government policy that controls the behavior of a private company? The word for that kind of policy is "regulation".
InsideOutSanta 1 days ago [-]
There is no policy.
sanderjd 21 hours ago [-]
Everything the government does is a policy. This is a policy. This is regulation. It's just a very bad policy, a very bad way to do regulation.
InsideOutSanta 9 hours ago [-]
I'm not sure whether there is a widely accepted definition of "policy" that matches yours, but if there is, it's not a helpful one. In general, a policy is a deliberate system of rules or guidelines.
sanderjd 4 hours ago [-]
Yes there is, and it is a helpful one. It is good and helpful to have words that just mean a simple thing. The things an administration does are the policy of that administration. That policy been be a deliberate system, or it can be haphazard and arbitrary. It is still their policy either way.
8note 1 days ago [-]
bad regulation is still regulation
InsideOutSanta 1 days ago [-]
> regulation /,rɛɡjʊ'leɪʃən/
>noun
>the act of controlling or directing according to rule
So where is the rule? There is no rule. There is just a random order. This is not regulation.
> Amazon removes guardrails from Fable, getting access to Mythos.
Amazon did not remove any "guardrails" from Fable. They created a fake, obviously insecure program. And apparently their prompt was exactly, "Fix this code." And Fable fixed the bugs.
This is something that even dinky local Chinese models running on a high-end gaming GPU can often do. Certainly Opus, GPT 5.5 and Gemini can all do this. And any high-end Chinese "near-frontier" model can do this, too.
But either (1) the administration is too clueless to know most models can do this, (2) Trump wants to be paid a bribe, (3) someone thinks Anthropic is "woke" and should therefore be destroyed by the power of the state, or possibly, if you're really cynical, (4) maybe the NSA SIGINT wants access to Mythos so they can break into everyone's computers, but they don't want you to have a model good enough to keep them out. Take your pick, I guess.
Anyway, apparently we don't do free markets or rule of law in the United States any more?
rdtsc 1 days ago [-]
Didn't their CEO argue how super powerful and dangerous their AI models are and government should be able to restricting usage
> "We're proposing stronger regulation of the technology, proposing giving the government the ability to, again, in a narrow way, block deployment of unsafe technology," he said.
The government: "We're restricting its usage to US citizens only"
Anthropic: "No, no, not like that! "
sanderjd 1 days ago [-]
I mean, yes, correct, literally "not like that" is the complaint. Government arbitrarily picking and choosing what's allowed in order to force everyone to curry favor is very different from an general well-documented regulatory framework. It is not weird for someone who favors the latter to call foul at the former.
rdtsc 1 days ago [-]
> Government arbitrarily picking and choosing
Their CEO was asking for it. Their whole marketing angle is their model is so powerful and dangerous.
Someone showed the government how the powerful and dangerous features can be unlocked with a 'fix bugs' prompt then it when and did exactly what their CEO asked for.
sanderjd 1 days ago [-]
And yet, "not like this" is still the correct conclusion to draw with respect to this kind of corrupt political favoritism.
rdtsc 1 days ago [-]
Where is the favoritism? Did other CEOs come out and say their models are crazy dangerous? Dario asked for this behavior and he got it. I guess he hoped to kneecap Deepseek or others. That's how companies operate, once they are big enough they want all the regulations because they know they can navigate them and others catching up may not be able to. Their own fear mongering around this demand backfired.
sanderjd 22 hours ago [-]
Yes, everyone in the space has front loaded the concerns about the power of this technology. Dario did not "ask for" singling out by a gangster administration. I probably agree with you about regulatory capture, but just "we don't like you, you haven't licked our boots enough, so we're going to screw with you" is a totally different thing. It's a different thing that is new to the modern history of the US. It's not new in general, lots of governments have been corrupt in this same way throughout history. But that's bad, and we shouldn't adopt that approach in the United States.
I agree that the fear mongering is bad, but that's a total pretext for what happened here.
khalic 1 days ago [-]
You guys are going to have to find another argument because this is just stupid. Being singled out doesn’t equate to regulation being applied.
The story is: anthropic refuses to give the US give an abliterated version of Claude for their weapons system, the US gov retaliates.
You’re rooting for the mass murderers, good job
rdtsc 1 days ago [-]
> You’re rooting for the mass murderers, good job
How did you figure that? Who are you talking about?
christkv 21 hours ago [-]
I hear this argument about the weapon a lot but it makes no sense. Sure for intelligence processing etc but what weapons system would an llm be a good match for.
hintymad 21 hours ago [-]
When I look at the timeline, it appears that they got exactly what they've been asking for. Or what did I miss?
April: Anthropic announces Claude Mythos. Citing safety concerns regarding the model's capabilities, they restrict access to a small group of testing partners.
May 15: Anthropic executive Chris Olah visits the Vatican to assist the Pope in publishing a 40,000-word manifesto, Magnifica Humanitas. Olah publicly states the model exhibits over 100 distinct emotional traits.
The pre-release: Anthropic briefly calls for a global halt to AI research.
The release: Anthropic launches Claude Fable 5, a consumer-facing version built on the Mythos architecture but with strict safety guardrails.
The post-release throttling: Users discover Fable 5's performance is being silently degraded on specific tasks. Anthropic acknowledges the throttling, citing the need to prevent foreign competitors from extracting synthetic training data. They promise future limitations will be transparent. Performance is notably restricted on biology and chemistry prompts.
The essay: CEO Dario publishes an essay advocating for strict government oversight. He proposes heavy regulation for any AI company exceeding $500 million in AI revenue or $1 billion in R&D spending.
The regulation: Two days later, the US government places export controls on both Mythos and Fable 5, classifying them alongside high-end semiconductor equipment. This immediately blocks Anthropic's own non-US employees from accessing the models.
The fallout: Amodei calls the export controls a "misunderstanding" that disrupts their global internal operations. Anthropic subsequently issues a statement noting that known jailbreaks for Fable 5 are also effective on OpenAI's GPT-5.5, highlighting the lack of equivalent regulatory action against their competitor.
mrbombastic 17 hours ago [-]
Presumably if they want ai regulation they want evenly applied regulation across the industry, not just for them.
Rendered at 18:19:48 GMT+0000 (Coordinated Universal Time) with Vercel.
Government was supposed to be high-trust, but Nixon weakened the trust, Reagan explicitly took aim at it and lots of people went along. Now the reactionaries run things. Cancer in the brain.
There's two parts to this. The first is that in most cases, the issues stem from lack of clarity in existing legislation, or contradictions thereof. At anytime, if Congress does not like the way the Supreme Court ruled, they can pass different legislation.
The second stems in part from the first. If what Congress passed is interpreted as being unconstitutional by the Supreme Court, the States can amend the constitution.
There is no case where the Supreme Court can (in principle) simply make their own rules*
The debate and controversy comes where the Supreme Court is seen as making their own rules (or arguably asked to) because the legislative process is deemed too cumbersome or disadvantageous to a party. In my opinion, this is where a lot of the difficulties lie, in that the Supreme Court is asked to rule on matters that really should be more clearly legislated. But the issues are seen as easier/quicker to address by convincing 9 justices rather than two legislative bodies and a chief executive. Naturally, this brings it's own consequences.
The Voting Rights Act might disagree on that.
> There is no case where the Supreme Court can (in principle) simply make their own rules.
With at least one significant exception.
https://en.wikipedia.org/wiki/Marbury_v._Madison
US Constitution Article III Section 2 Clause 2 which says the supreme court's appellate jurisdiction is subject to "Exceptions, and under such Regulations as the Congress shall make." So Congress could, for example, require the supreme court to hear every case appealed to them instead of letting the justices pick their cases. Or it could require judges to recuse themselves if there is a conflict of interest*. Or it could forbid the court hearing cases where an en banc appeals court was unanimous.
* which could have more teeth than the self-imposed ethics rules for conflict of interest that the court has now.
The US desires to show its own citizens and the world how painful it is not to submit to the US government. They're leveraging everything they've got.
The only rational response when faced with this sort of behavior is to reduce dependence on and exposure to the US as much as possible. Develop national technology. Dedollarize the economy. Invest in and use open weight models.
Institutions are not built around zero trust. Because when we assume attackers are identifiable, and there is some level of integrity, that makes things a lot simpler and cheaper.
I will not expect my pediatrician to want to kill my child and they certainly could, but it would be completely bonkers to try to guard against that.
The level of damage that a single really motivated, disgruntled person with a lot of power can do in most settings is intense. The cost inflicted if we tried to build guardrails against all of it is insane, and it probably would never work.
Traditionally this was applied to people in power with the power of a good old fashioned lynch mob. If you obtain great power and use it to start hurting large numbers of people, the citizenry at large would drag that person out into the street and behead or beat them to death.
It worked pretty well for all societies throughout history, including the West. If a person becomes a danger to society, they should be removed from society with proportional prejudice. That's the guardrail society has always had.
Problem is now that those in power have managed to convince large fractions of the population that it's wrong to say mean things about sociopathic mass murderers and child rapists or the guy who is actively trying to kill you personally.
Unfortunately we don't seem to have the ability to do the same for corporations. If a person starts killing or robbing people, they get arrested and are removed from society. If a corporation starts killing or robbing people, they maybe get a strongly worded letter, maybe have their CEO say a few words in front of Congress, and maybe get a token fine amounting to 0.001% of their revenue (which they will appeal for ten years). But for whatever reason, we don't seem to have the guts to remove a badly behaving corporation from society--no matter what heinous things they are doing, no matter what laws they are breaking.
Actually, civil societies granted a monopoly on violence to the government which vests that in police and military. It’s a far better system than random justice because it’s predictable and you know upfront what’ll trigger it.
That system had its flaws, but despite that it was working pretty well until those flaws were found and exploited allowing it to be captured.
The U.S. is either harming itself erratically, or systematically enforcing control on US based businesses in a way that is historically more synonymous with, say, China.
A lot of us worry it is the former. I wonder if it's better or worse if it is the latter. The US, having seen that Europe and the world seem to tolerate interference by CPC based on opinions expressed by Alibaba or Chinese car companies, may have decided that it's fair game.
[1] https://www.youtube.com/watch?v=eqoX9OIR-DI
I think this can be said in many contexts these days.
For the most part no. The Administrative Procedures Act outlines a fairly robust process for rulemaking (ie regulations must go through a notice period, must be justified etc) and adjudication (ie per-person decisions must have notice, ability to have counsel, be appealable etc).
However there's also lots of authority given to the president (and executive office of the president) to which the APA doesn't apply. This includes powers the constitution gives to the president directly and (mostly emergency) powers congress gives to the president. Sadly political processes and supreme court cases have hugely increases what can be done with this*.
Moreover, it's been a frequent pattern (over many decades, but especially under Trump) for the president to simply nakedly overstep their authority. Sometimes there's no push back (see: 9/11); sometimes the damage is done even if it's eventually overruled. Never is there any consequences though, so the pattern repeats.
* E.g. The president's control of the military is supposed to be checked by congress deciding when to declare war. However the war powers act has in practice played out to allow the president to unilaterally start wars, especially as the supreme court has stepped in to make the war powers act less effective.
Is anyone still assuming the administration's actions have anything to do with US national interests?
Anthropic CEO https://abc7.com/post/anthropic-ceo-dario-amodei-calls-stron...
https://abc7.com/post/anthropic-ceo-dario-amodei-calls-stron...
(whether Mythos is actually that dangerous is beside the point; considering that Anthropic claims that it is, it makes sense to regulate it)
Goverment: ...
Anthropic CEO: NO, NO, NO NOT LIKE THIS.
You can come up with and advocate for a policy to be evenly applied to all the billionaires whose "misfortunes are our freedom", instead of supporting this corrupt favoritism.
He told the world about 20 months ago that all software developers would be gone in 12 months. Why has he not fired all Anthropic Devs yet?
What you're advocating for here is banana republic governance. It's bad.
I’ll make them argue that when you get Sam Altman to admit he lost the lead to anthropic
Perhaps, ITAR and the like isn't my legal specialty.
Isn't on a company's CEO to make sure everything is copacetic with its public and private partners?
This is an incredibly dumb way to judge a government's behavior.
There is also a case that it violates due process because anthropic wasn’t provided adequate notice before enforcement or an opportunity to challenge, respond, or take steps to comply.
There is latitude for compelling state interests and national security is one with the widest latitude and potentially one that counts can’t review. But the court of public opinion may draw the inference that this is direct retaliation for refusing to make the models available for autonomous weapons and spying.
Instead, one morning the administration just woke up, grabbed the nearest hammer, and went smash, like it usually does.
I'm sure it says something that you think makes this a slam-dunk rebuttal, but you should fix your link first.
Edit: Saw the link. I was right. Nothing in that link makes a difference.
Any government work can be done via MOU with a US subsidiary staffed by Americans.
This kind of capricious, unexplained control is bullsh*t.
I’m not saying that I want companies to have to go offshore or that that would be a good thing. Just that you’ve got no leverage if your corporate structure can be destroyed on a whim. This goes for any company reliant on a whimsical executive branch.
While there could still be fights over the technology and the company, a tech provider would still be able to serve other customers and have more leverage.
And drugs, lots of drugs.
Like where? How about Cuba? Where Guantanamo is, where we put terrorists because we don't want them to be subject to US laws (or more specifically, we don't want ourselves to be subject to those laws when we are dealing with said terrorists). How about no.
International law has only as much power as the biggest military willing to enforce it. Hiding on some little island is not at all a good strategy for trying to evade the US government.
That would seem to me as grounds for piercing the corporate veil and holding the people behind it directly responsible.
High treason or some similar charge. Capital punishment not out of consideration.
This "across the border" strategy will not work.
Seagate Singapore was hit with civil penalties and settlements for shipments to Huawei under the Foreign Direct Product Rule.
ZTE, a Chinese company, was criminally prosecuted and settled with U.S. DOJ/BIS in 2017 for violating U.S. export controls and sanctions.
I hope those who represent technology in government, especially the AI head David Sacks are giving this due consideration.
It seems to be observably true that expressing opinions critical of the administration, even about one's own business, leads to further harm from government right now in the U.S. in a country which is supposed to have strong freedom of speech rights.
It's obviously unfortunate that this is the case.
It's also a shame that this is happening when the folks regulating AI have a tech industry background.
The administration seems to be clearly signaling "stop resisting". I am not sure how you read me as saying that.
I can’t rely on using a technology that the US administration can ban at will.
IMO without getting into personal thoughts about how capable the current US administration is, last Friday move sent a very powerful signal to the industry.
Also I don’t think China releasing so many good models, capable to compete with Opus 4.8 and GPT 5.5, all at once, is a coincidence.
It seems to me that in the case of AI (as with many other modern technologies), you rely on vendor/creator support and updates to stay relevant, so the ‘next’ model matters more than the current one, and we have no idea whose next model will be open (and whose won’t).
I suspect AI will be somewhat similar where even if the linear scaling laws continue to hold the practical utility of a model flattens for almost all conceivable use cases.
In some ways I already feel this has begun to happen. The marginal utility of opus class models and fable has in my perception begun to flatten. While I can tell the differences they aren’t earth shattering. I could continue to use the present models for the rest of my life and be ludicrously more productive simply by adapting within their constraints through ever more sophisticated applications.
What holds back the open weights IMO is hardware scaling and industrial production. As the enormous transfer of wealth in debt and equity markets unfolds with semiconductor and adjacent companies and the corresponding capital investments are made, and the eventual bubble pop leading to over capacity and market flooding, as well as advances in technology, math, techniques, and efficiencies, will make very large open weight models more directly attainable. This will also lead to chimera models that MOE very large models to get very close to the 1-2T parameter dense models, at which point I suspect utility for almost all uses is nearly fully saturated.
There will be areas where more capable models are needed but they will be frontier models on frontier problems. This, IMO, is inevitable, and without some criminalization of weights (see the attempts to criminalize encryption algorithms in the 20th century and all the wonderful tshirts that emerged). It’ll be harder to print a trillion parameter model on a shirt but I’m sure someone will try, as will governments try to keep us in our boxes slaving for food coupons and basic rights like health care.
Why not both?
That seems the crux of the state we're currently in; what daylight there was between the two is quickly fading.
I didn't realize I could download a Shanghai apartment.
I wholeheartedly agree with what you said about China.
But I can’t shrug off the fact that fable was taken down within minutes for reasons that are childish and petty.
I am sorry but I can’t use any US AI if I don’t have the guarantee that I will be able to use it tomorrow.
And Trump showed us he is willing to take it out whenever he wants.
An opensource model on the contrary, I can host myself, or use a miriad of providers, mostly non chinese.
To be fair this is every commercial model. We have already seen GHCP increase prices by anywhere from 10-100x (depending on usage). And old models get retired all the time. While these are not exactly the same as a cutting edge model being shut down, increasing prices a super high amount leads to effectively the same outcome.
Yes, the actions of this administration on Friday should have sent shockwaves through the market - a market that's currently "high on AI". How do you get a return on all of that AI investment if the administration can jump in at any time and say "Nope, you can't use this very advanced model!"? (the Iran "deal" over the weekend, I think helped cushion that blow, but eventually it's going to sink in)
The latest open models are so good it’s worth the 6-8 months delayed capabilities. At least for coding
Fable was around 10x GPT5 pricing and 100x Chinese models pricing, was it really 100x better? I Don't think so.
If you want a personal story, I just solved a complicated coding problem with Kimi 2.7 that GPT 5.4 failed with.
And you think China will not do the same thing if their models ever become genuinely frontier-level?
What’s intolerable is having a tool that’s subject to this risk.
So open models it is
The impact will last long, long, after this administration is gone. Trump will have his legacy.
I would hope they return to their home countries with their expertise and start or join new competing frontier labs, similar to how Taiwan's homegrown semiconductor industry arose from US companies enforcing a bamboo ceiling on their Asian engineers. Taiwan was able to repatriate their nationals and make incredible compounding leaps in the semiconductor industry, to the loss and chagrin of the US. [1]
[1] https://www.npr.org/2022/10/07/1127595393/taiwan-miracle-sem...
We're all dependent on Taiwan but others also depend on ASML which is located in Europe. Available AI services are good enough, and we sit and watch while US leadership and population is beta testing AI mass surveillance for us.
In terms of AI-based warfare Ukraine is very much ahead and they are also European for which we are very thankful.
So feel free to keep your AI scene and AI hustlers in the US, we don't need them.
Does this actually matter as much as some people think? Probably not because LLMs won’t lead to AGI, but it is indicative of a broader problem
Expensive publicly funded European education for 25 years -> high paid job in the US.
Costs are all paid by Europe, tax revenue and job creation all accrue to the US.
If we’re not careful it ends up being a straight transfer of value from ordinary European taxpayers to the US government which seems… suboptimal.
If the US can short circuit that process by shooting themselves in the foot, Europe will benefit
It’s what China does (except for poaching people from Taiwan), but they have 1bn people so for them it makes sense
Note that the Hornet drones which are now devastating Russian logistics behind the lines are US manufactured.
Here right from a Ukrainian military server.
"it is not noteworthy"
From a peaceful office in Germany, it might not be. For a country fighting for survival, the perspective is different. I don't think you could persuade them to give up Hornets right now, in the name of abstractly getting back at the US in general. (With whom they have a lot of bad blood!)
There is no way any employees at anthropic are this dumb.
I used Fable last week to scan my own work for security errors (and readability issues), I wasn't thinking about the ban on infosec use I just wanted the most well trained output.
One project in the monorepo triggered a Opus fallback and infosec/biology warning.
The other project scanned fine and showed a bunch of potential exploits.
- They're obviously being targeted politically because they refuse to kiss the ring, vibes, whatever you want to call it.
- They're also justifiably being scrutinized because they just spent like 3 months telling everyone that Mythos is a nuclear bomb and telling the government to fuck off as they drip fed access to a bunch of random corporations.
But in cases like this, the pretext shouldn't be taken too seriously. Because they would have just found another one. It isn't actually particularly important that it is plausible, it's more like a happy coincidence.
I don't understand why their marketing department/execs can't see the conflict between claiming AI is going to take all jobs, that the model is super dangerous and AI hate among the general public, increased governmental oversight.
Anthropic is guilty of the latter but the former applies to most of these AI companies.
So their plan is:
1. We can't stop other people from building something dangerous.
2. But we can get there first.
3. If we build it, it has maybe a 15% chance of killing everyone alive. (I think that's a number I've seen Dario use before, but I may be wrong.) If OpenAI or China build it, the odds would be worse.
Obviously, if Anthropic is actually correct about (1) and (3), then nobody should allowed to build frontier AI.
People find it really hard to believe that (a) anyone believes in the possibility of dangerous AI in our lifetimes, and (b) that someone could believe what Anthropic seems to believe and then still go ahead and gamble with everyone's lives anyway.
This has some insight in my many cyber security professionals want Fable/Mythos open.
This particular executive loves money, praise, and submission. When Anthropic submits (eg agreeing to whatever the DoD demands), issues public praise and makes appropriate donations (eg ballroom, memecoin, etc), then they can do as they please.
Students of history will find this new MO very familiar and very depressing.
Anthropic themselves refused to allow the Pentagon use of its models for any legal purpose - instead the wanted a subset picked by Anthropic. It's ironic that you're accusing the government of ignoring the law when Anthropic openly wishes to set it's own private laws.
They're not simply petty; instead, they're vindictive, retaliatory, and vengeful.
Elon vs openai
Trump vs elon
Bezos vs anthropic
When movie?
Anthropic reaping: Hey wait a second you weren't supposed to take that seriously it was just marketing :(
If the Trump admin is also willing to apply the same scrutiny to GPT-5.6 and other Fable-level models, I think it is a good thing. But given the admin's history with Anthropic (such as declaring it a supply chain risk while ignoring Chinese labs), there is some smell of targeting.
So at some point - don't say your model is capable of hacking even if it is and you are clean.
Now they want them back on?
AI for me and not for thee?
Or just another opportunity for a petulant administration to target Anthropic for not doing what they wanted the first time?
100%, I mean we already have Mythos vs Fable
How is this “targeting”? It’s literally what was requested.
Based on the reasoning for blocking Fable, every current model should be blocked. GPT 5.5 is similarly strong and has fewer guardrails, for example.
I've heard others call it a marketing ploy that AI companies keep coming back to-- the idea that "it's soooo dangerous, we have to be careful".
Yes, most likely, but not in this form, obviously. They want open weight models regulated for regulatory capture, and I'd assume they want an actually documented framework applied equally across all labs.
If GPT5.5 has the same capabilities Fable did, then for consistency sake, it should have also been subject to this ban.
Regulate or not regulate, but the government should not pick winners and losers.
This doesn't seem like an accurate description to me. I think something like "Amazon demonstrates a jailbreak of one class of Fable guardrails" would be a more accurate description.
It doesn't even really mess up your narrative to state it accurately, but your choice of a more hyperbolic statement brings into question the good faith of the narrative you're painting.
https://news.ycombinator.com/item?id=48552687
Anthropic themselves specifically called them safeguards. [1]
"When Fable’s classifiers detect a request related to cybersecurity, biology and chemistry, or distillation, the response is automatically handled by Claude Opus 4.8 instead"
This is exactly what was bypassed. They got Fable to work on security topics.
[1] https://www.anthropic.com/news/claude-fable-5-mythos-5
This is Anthropic's own claim. They were very specific. Have you read their own claims?
"When Fable’s classifiers detect a request related to cybersecurity, biology and chemistry, or distillation, the response is automatically handled by Claude Opus 4.8 instead. Users will be informed whenever this occurs."
Asking Fable to fix bugs in a code base is not "a request related to cybersecurity." When Fable was asked to fix bugs and then proceeded to fix bugs, that was not "removing guardrails". Fable did exactly what it should have done. Claiming otherwise makes absolutely no sense at all.
Anthropic specifically stated that ANY security requests should be shunted to Opus 4.8. This was bypassable.
I don't see what your confusion here is. Fable was prevented from working on any security tasks. A significant amount of people, myself included, witnessed Fable refusing to harden code as a result. Bypassing that is a bypass of guardrails.
Your assertion that working on security is not working on security because you used misdirection is of course, preposterous.
You wouldn't be making the same claim if Fable refused to work on chemical weapons research but happily proceeded to do so if you claimed it was for eradicating pests.
Asking a model to fix bugs is neither misdirection nor a security request.
> I don't see what your confusion here is.
That's because I'm not confused :-)
> Fable was prevented from working on any security tasks
I don't think that's true based on what Anthropic said, and I also don't think it can be true.
What do you propose Fable's behavior should be if you ask it to fix bugs, and it encounters a security issue? I'm assuming your solution is that when you ask Fable to "fix bugs," and it encounters a bug that could be exploited as a security vulnerability, it should fall back to 4.8. But that doesn't solve the problem, because as a user, I can now see where that occurred, so I still know where the vulnerability is. That's not substantially different from the current outcome, where it just fixes the bug.
It would also mean that Fable could barely make it through any code review without falling back to 4.8, because almost any non-trivial code base has aspects that could be interpreted as security vulnerabilities.
The alternative would be for the model to use its hidden thinking to decide not to fix the bug, but that seems even worse.
>noun
>the act of controlling or directing according to rule
So where is the rule? There is no rule. There is just a random order. This is not regulation.
https://news.ycombinator.com/item?id=48552687
Amazon did not remove any "guardrails" from Fable. They created a fake, obviously insecure program. And apparently their prompt was exactly, "Fix this code." And Fable fixed the bugs.
This is something that even dinky local Chinese models running on a high-end gaming GPU can often do. Certainly Opus, GPT 5.5 and Gemini can all do this. And any high-end Chinese "near-frontier" model can do this, too.
But either (1) the administration is too clueless to know most models can do this, (2) Trump wants to be paid a bribe, (3) someone thinks Anthropic is "woke" and should therefore be destroyed by the power of the state, or possibly, if you're really cynical, (4) maybe the NSA SIGINT wants access to Mythos so they can break into everyone's computers, but they don't want you to have a model good enough to keep them out. Take your pick, I guess.
Anyway, apparently we don't do free markets or rule of law in the United States any more?
https://abc7.com/post/anthropic-ceo-dario-amodei-calls-stron...
> "We're proposing stronger regulation of the technology, proposing giving the government the ability to, again, in a narrow way, block deployment of unsafe technology," he said.
The government: "We're restricting its usage to US citizens only"
Anthropic: "No, no, not like that! "
Their CEO was asking for it. Their whole marketing angle is their model is so powerful and dangerous.
Someone showed the government how the powerful and dangerous features can be unlocked with a 'fix bugs' prompt then it when and did exactly what their CEO asked for.
I agree that the fear mongering is bad, but that's a total pretext for what happened here.
The story is: anthropic refuses to give the US give an abliterated version of Claude for their weapons system, the US gov retaliates.
You’re rooting for the mass murderers, good job
How did you figure that? Who are you talking about?
April: Anthropic announces Claude Mythos. Citing safety concerns regarding the model's capabilities, they restrict access to a small group of testing partners.
May 15: Anthropic executive Chris Olah visits the Vatican to assist the Pope in publishing a 40,000-word manifesto, Magnifica Humanitas. Olah publicly states the model exhibits over 100 distinct emotional traits.
The pre-release: Anthropic briefly calls for a global halt to AI research.
The release: Anthropic launches Claude Fable 5, a consumer-facing version built on the Mythos architecture but with strict safety guardrails.
The post-release throttling: Users discover Fable 5's performance is being silently degraded on specific tasks. Anthropic acknowledges the throttling, citing the need to prevent foreign competitors from extracting synthetic training data. They promise future limitations will be transparent. Performance is notably restricted on biology and chemistry prompts.
The essay: CEO Dario publishes an essay advocating for strict government oversight. He proposes heavy regulation for any AI company exceeding $500 million in AI revenue or $1 billion in R&D spending.
The regulation: Two days later, the US government places export controls on both Mythos and Fable 5, classifying them alongside high-end semiconductor equipment. This immediately blocks Anthropic's own non-US employees from accessing the models.
The fallout: Amodei calls the export controls a "misunderstanding" that disrupts their global internal operations. Anthropic subsequently issues a statement noting that known jailbreaks for Fable 5 are also effective on OpenAI's GPT-5.5, highlighting the lack of equivalent regulatory action against their competitor.