Notes ·
What Does "Unlawful" Mean If the Damage Happens First?
Joshua MorrisI keep coming back to a practical question.
What does it mean for the government to do something unlawful if the consequences happen immediately and the legal determination comes later?
This isn't really about defending Anthropic.
It is about whether the law can meaningfully constrain government action when modern political and economic consequences can happen almost instantly.
In February, the administration directed federal agencies to stop using Anthropic's technology, and Defense Secretary Pete Hegseth designated the company a supply-chain risk.
That designation potentially reached much further than the Pentagon itself. Companies doing business with the military were also told they could not conduct commercial activity with Anthropic.
Anthropic sued.
Now U.S. District Judge Rita Lin has ruled that the government's actions were unlawful.
Her ruling is unusually direct.
The court found that the government retaliated against Anthropic for constitutionally protected speech in violation of the First Amendment.
It found that Anthropic was denied the process required by the Fifth Amendment before being deprived of important business interests.
And it found that the supply-chain-risk designation violated the statute governing that authority and was arbitrary and capricious.
The court's conclusion wasn't merely that someone filled out the wrong form.
The judge found that the government had used these powers to punish Anthropic for publicly disagreeing with it.
That should matter.
The harder question is how much.
The legal system actually moved fairly quickly here
My first instinct was that this is another example of the law moving at horse-and-carriage speed while political decisions happen at internet speed.
That isn't entirely fair in this case.
Anthropic filed its lawsuit on March 9.
On March 26, Judge Lin issued a preliminary injunction blocking enforcement of the challenged directives while the litigation continued.
Seventeen days is remarkably fast for federal litigation.
The injunction stopped the government from enforcing the supply-chain designation, implementing the broader directives, or issuing additional guidance designed to carry them out.
The court also required the government to report back on its compliance.
So emergency mechanisms do exist.
Federal courts can issue temporary restraining orders and preliminary injunctions.
Appeals courts can issue stays.
And, at the other end of the process, the Supreme Court has an emergency docket capable of acting in days rather than months.
The legal system is not completely incapable of moving quickly.
This case is actually evidence that it can.
But speed is only part of the problem.
You can't unring every bell
Imagine that the government publicly declares a company a national-security supply-chain risk.
Customers reconsider contracts.
Partners hesitate.
Competitors exploit the uncertainty.
Employees wonder what is happening.
Investors price in political risk.
Government agencies begin planning migrations.
News stories repeat the allegation.
Then, weeks later, a judge says the designation was unlawful and blocks it.
The injunction matters.
But it cannot make those weeks unhappen.
That is the asymmetry I find troubling.
An executive action can create consequences immediately.
Judicial review is necessarily reactive.
Someone has to be harmed enough to challenge the action, hire lawyers, file a case, establish jurisdiction, present evidence and persuade a judge that emergency intervention is justified.
Even when that happens quickly, the unlawful action gets a head start.
The final ruling arrives months later.
And this case moved faster than many.
"Illegal" and "consequential" are different questions
I think I have been conflating two questions.
The first is:
Was the government's action lawful?
The courts are designed to answer that.
The second is:
What happens to the government when the answer is no?
That is much less straightforward.
Anthropic's lawsuit primarily sought declaratory and injunctive relief.
In other words:
Declare these actions unlawful.
Vacate them.
Stop enforcing them.
Tell government agencies to disregard them.
Those are important remedies.
But they mostly put a stop to future harm.
They do not necessarily compensate for everything that happened while the policy was operating.
That creates an incentive problem.
If an official knows that the worst likely outcome of pushing the boundaries of their authority is that a court eventually says "stop doing that," there may not be much institutional cost to trying.
The policy may already have accomplished some of what it was intended to accomplish.
I don't think that problem is unique to this administration.
The expansion of executive power has been an issue for a long time.
I remember becoming uncomfortable with it during the George W. Bush administration, particularly around national security.
Then administrations I was more sympathetic to inherited those powers and sometimes expanded them.
That made the underlying problem harder to ignore.
It is easy to tolerate additional executive authority when the person exercising it is someone you trust.
The precedent remains when the office changes hands.
The presidency has accumulated an enormous amount of practical power over a very long period.
Eventually someone you don't trust inherits everything the people before them normalized.
Faster courts don't completely solve this
It is tempting to say the answer is simply faster judicial review.
Sometimes it probably is.
If a government action has immediate national consequences, courts should have mechanisms for reviewing it quickly.
They already have some.
But this Anthropic case makes me think speed is not the hardest part.
Anthropic got preliminary relief in 17 days.
Suppose it had taken seven.
The original declaration still happened.
The headlines still happened.
The business uncertainty still happened.
And an administration willing to aggressively test the boundaries of its authority might make roughly the same calculation.
The deeper question is how to design remedies so that unlawful government action is expensive enough to discourage without making public officials afraid to make legitimate decisions.
That is difficult.
Too little accountability and there is an incentive to act first and let courts sort it out later.
Too much personal liability and officials become afraid to exercise authority at all.
Somewhere between those extremes has to be a system where ignoring statutory and constitutional limits carries a meaningful institutional consequence.
I don't know what the right mechanism is.
Damages?
Automatic expedited review?
Stronger congressional oversight?
More meaningful consequences for agencies that repeatedly exceed their authority?
Some combination of those?
But "a judge eventually said it was illegal" doesn't feel like a complete measure of whether the rule of law worked.
The remedy may matter more than the ruling
There is an encouraging part of this story.
The court did intervene.
The preliminary injunction prevented the government from continuing to enforce the challenged actions while the case proceeded.
And now the court has ruled on the merits rather than merely deciding that Anthropic was likely to succeed.
That is the legal system working.
But the case also illustrates its limits.
Government can act in hours.
Markets can react in seconds.
Reputations can be damaged by a single post.
The courts arrive afterward and try to reconstruct what should have happened.
Maybe the important measure of the rule of law isn't simply whether a court can eventually declare something unlawful.
It is whether the system can stop the unlawful action before the consequences become irreversible, and whether there is enough consequence afterward that doing it again isn't an attractive strategy.
Otherwise "unlawful" risks becoming a very important word with surprisingly little deterrent effect.