Constrained Chaos, Licensed Trust, and Harmful Advice

The law is what it is. Sometimes it even has an internal logic that we can follow, even when we are uneasy about where it leads.

Why This Post Exists

In _Chiles v. Salazar_, the Supreme Court held that Colorado’s ban on conversion therapy, as applied to a counselor’s talk therapy, regulated speech based on viewpoint and therefore required stronger First Amendment scrutiny than the lower courts had applied. That holding is narrower than “health professionals can now say anything they want,” but broad enough to reopen an old and uncomfortable question: where does professional duty end and protected speech begin?

The Question

The question I want to sit with is not whether the Court can articulate a theory. It can. The question is this:

If a profession is licensed because clients are vulnerable and rely on expertise, why should the First Amendment make it harder to stop advice that is predictably harmful?

It gets even stranger when we compare health advice to financial advice. In the investment world, the law is quite comfortable imposing a best-interest framework. Since June 2019, the SEC’s Regulation Best Interest has required broker-dealers making recommendations to retail customers to act in the customer’s best interest, and the SEC has also reaffirmed that investment advisers owe a fiduciary duty of care and loyalty that requires serving the client’s best interests rather than subordinating them to the adviser’s own.

So here is the puzzle: in some settings, we seem more comfortable demanding protection from bad advice about our money than from bad advice about our bodies, identities, or futures. That is not a slogan. It is a real asymmetry in how the law sorts professional relationships.

Why this feels inconsistent

In ordinary life, the justification for licensing is straightforward. We license professions when people are vulnerable, when expertise matters, and when bad advice can cause serious harm. We do not do it for decoration.

A licensed counselor, physician, nurse, lawyer, or adviser is not just another citizen with an opinion. This is someone the public is taught to trust, someone invited into consequential decisions because the state has recognized them as qualified. That is why it feels odd when the law says, in effect, “this professional relationship matters enough to regulate, but not enough to restrain some categories of harmful guidance.” The Court’s reasoning in _Chiles_ rests on free-speech doctrine, especially concern about viewpoint discrimination, but the client does not experience the encounter as abstract doctrine. The client experiences it as advice delivered inside a licensed, unequal relationship.

There is another layer. The major professional bodies are not neutral on the underlying practice. They have described conversion efforts as ineffective and harmful. So the legal protection here does not arise because the practice has become medically sound. It arises because speech doctrine is doing heavy constitutional work.

Why would we be more risk-averse with money than with health?

When money is involved, the law often uses concepts like fiduciary duty, conflict of interest, disclosure, suitability, and best interest. We expect regulation because we assume the client can be misled, steered, or exploited. We understand that markets create incentives that can distort advice, so we build legal structures to constrain that distortion. We do this in a domain where bad advice can absolutely ruin lives, but usually through balance sheets, retirement accounts, and debt exposure.

But health is at least as vulnerable to distortion, and arguably more so. People seeking health care are often frightened, dependent, inexperienced, and operating under time pressure. They may be teenagers. They may be pregnant. They may be depressed. They may be sitting across from someone who speaks with the authority of training, licensing, and institutional legitimacy. If there is any domain where “best interest” feels morally intuitive, it is here.

And yet the legal system is far less consistent in how it handles that intuition. That is part of what makes this so jarring: the industry most associated with bankruptcy, financial collapse, and predatory incentives is regulated with remarkable comfort, while health advice that can alter identity, bodily autonomy, or long-term psychological well-being is treated as if the danger of state overreach is easier to see than the danger inside the professional encounter itself.

Part of the explanation is doctrinal. Financial advice often lives inside frameworks the law is already comfortable regulating as fiduciary or commercial relationships. Speech in the clinical setting, especially when the “treatment” is literally conversation, can look more like core expression to courts worried about censorship. That distinction may explain the doctrine. It does not dissolve the moral tension.

This is not only an LGBT+ issue

The inconsistency becomes even more visible when we move into reproductive care.

In some contexts, the government has tried to prevent clinicians or clinics from steering patients toward abortion or even referring for it. In others, the government or states have required clinicians to deliver specific counseling scripts before abortion, including waiting-period counseling. And in another line of cases, the courts have emphasized strong First Amendment protection against compelled speech.

In one setting, the state says, “you may not say this.” In another, it says, “you must say that.” In another, it says, “because this is speech, the Constitution limits our ability to require or forbid it.” The result is not a tidy principle consistently applied across health care. The result is a patchwork where the line between patient protection and ideological control can shift depending on the topic, the jurisdiction, and the doctrinal lens being used.

That is exactly the sort of inconsistency that makes people suspicious, not because every actor is hypocritical, but because the system is trying to satisfy multiple values at once: autonomy, safety, pluralism, anti-censorship, professional regulation, democratic control, individual conscience. All of them matter. They do not fit together cleanly.

The strongest objection

There is, however, a serious objection that any argument like mine has to face directly.

The reason courts worry about regulating professional speech is not theoretical. Governments do abuse licensing regimes. States do try to smuggle ideology into professional rules. Officials do use the language of safety, ethics, and public welfare to suppress disfavored viewpoints. A doctrine that gives government broad power to decide which advice is too harmful to utter can become a doctrine that licenses official orthodoxy.

That is the best free-speech argument here, and it should not be waved away.

But admitting that danger does not end the inquiry. It simply means the question cannot be “regulate everything harmful” or “treat all professional advice as ordinary protected opinion.” The real question is what limiting principle protects clients without handing the state a blank check.

The inconsistency we should admit out loud

We seem to oscillate among at least three different instincts:

1. Professionals should be free to speak honestly, even when the state dislikes the viewpoint.
2. Professionals should be constrained from giving advice that is predictably harmful.
3. Professionals can be required to say certain things when the state believes patients need that information.

Each instinct has some logic. Each can be defended. But they do not always coexist gracefully.

The friction is not a bug in public debate. It may be a feature of constitutional democracy. We live in a system built from constraints that do not fully reconcile. We protect speech because governments are dangerous. We license professions because expertise matters. We regulate advice because vulnerable people can be harmed. We fear paternalism, but we also fear abandonment. That is the knot.

What I think is worth asking

Not “who is evil?” Not “which side is stupid?” Not even “what is the one right answer?”

The better question is whether we have been sufficiently honest about how selective our principles can become.

If we are willing to require best-interest behavior when someone is advising us about retirement, why are we so much less certain when someone is advising us about identity, pregnancy, mental health, or bodily autonomy?

If speech inside a licensed relationship is uniquely protected, when exactly does professional duty bite hard enough to protect the vulnerable?

If compelled speech is constitutionally suspect, why do we tolerate some mandated health scripts more comfortably than others?

If harmful advice is dangerous, are we applying that concern consistently—or only where it aligns with our preexisting moral intuitions?

Those are the actual terrain.

If I think something is missing from the debate, it is this: we need a clearer test for when licensed professional speech becomes regulable conduct. Not a slogan. A test. One possible starting point would ask four questions at once:

1. Is the speaker acting inside a licensed professional relationship?
2. Is the client meaningfully vulnerable or dependent?
3. Is the guidance contrary to the prevailing standard of care or professional consensus?
4. Is the state targeting demonstrable harm rather than disfavored viewpoint alone?

That test would not solve everything. But it would force us to say what we are protecting, what we are restricting, and why.

A note for the age of AI

This is not a post about AI, but in today's world, everything seems to be about AI.

We keep asking GenAI systems to give robust, context-sensitive, fair, safe advice in domains where human law, ethics, medicine, and politics remain deeply unresolved. That is not a reason for despair. It is a reason for humility.

If our own institutions cannot produce a stable theory of when speech becomes harmful professional conduct, or when autonomy yields to protection, or when neutrality gives way to mandated disclosure, then we should not be surprised that AI systems inherit some of that ambiguity. They are learning atop a civilization that often reasons by layered compromise rather than elegant consistency.

Constrained systems produce strange forms of order. Our constitutional life has always been a kind of constrained chaos—messy, adversarial, unstable, and yet somehow productive. The hope is not that we eliminate the chaos. The hope is that we become more honest about it, more careful with the vulnerable inside it, and more consistent in the principles we claim to hold.

That would be a good start.