I’ve crossed swords with Nobel Prize-winning economist Paul Krugman on health care. But in his latest Substack post, we are largely singing from the same sheet music (forgive the lengthy excerpts; I want to save all of this lest he delete it):

much of U.S. health care is already socialist. In fact, the government’s role is so large that U.S. health care is better described as partially privatized socialism than as anything resembling a free market…

government programs pay more total dollars for medical bills than the private insurers do…

Read the rest of this post →

A recent conversation I had with Claude should be mandatory reading for all policy wonks, reporters, and anyone who is either considering entering those professions or consuming their output.

In a conversation about a particular regulation, I noticed in Claude’s answers a pattern I have seen elsewhere, especially in health policy reporting:

On June 26, the United Nations Office on Drugs and Crime (UNODC) released its 2026 World Drug Report. Drug-war hardliners who (wrongly) credit recent overdose drops to tougher enforcement won’t like this report, which the agency summarized in a press release.

The latest report concludes that global illicit drug markets continue to adapt and expand. Traffickers increasingly exploit new technologies, shifting trade routes, geopolitical instability, and a growing array of synthetic drugs to evade law enforcement and reach new markets. The report estimates that 331 million people used illicit drugs in 2024 and warns that drug markets have become more diverse and resilient, increasingly driven by synthetic substances.

The report’s findings corroborate several observations that underlie my long-standing thesis:

• Drug markets adapt rather than disappear. 

The report emphasizes that traffickers rapidly adjust to enforcement pressure by shifting production methods, trafficking routes, and distribution techniques. As I wrote in testimony for the Senate Judiciary Committee in 2025, “as the US has pressured China to curtail precursor production, precursors are now coming from labs in India, Myanmar, other parts of Southeast Asia, and even Canada.” The press release states: 

The 2022 drug ban in Afghanistan has continued to severely constrain the illicit production of opium and heroin. Although production in Myanmar rose from 420 tons in 2021 to over 1,000 in 2025, the increase in the country (together with quantities produced in other countries monitored by UNODC, i.e., Laos and Mexico) does not offset the declines in Afghanistan, which in 2022 produced more than 6,000 tons of opium. 

The increasing availability of novel synthetic opioids such as fentanyls, nitazenes and orphines on the market suggests that traffickers are searching for alternatives to heroin. A turn away from plant-based opiates toward synthetics could cause a permanent shift in the global opioid market, with ramifications on how these drugs are used and the harms therein. 

• Synthetic drugs become increasingly attractive under prohibition. 

Synthetic substances are often easier to manufacture, require less agricultural land, are more concentrated, and are easier to conceal and transport than plant-based drugs. These characteristics make them especially well-suited to illicit markets. Fentanyl was among the first synthetic opioids to enter the underground market, but toxicology labs are reporting an increasing number of overdoses involving chemically unrelated synthetics that require different precursor ingredients, such as nitazenes and orphines like cyclorphine.

• Innovation is relentless. 

The report describes an “unprecedented spike” in novel psychoactive substances, underscoring that producers continually develop new compounds in response to market pressures and regulatory controls. Again, from the press release:

Illicit drug manufacturers continue to invent new synthetic drugs in attempts to skirt regulations and avoid detection, with five times more drug types found in seizures in 2024 than before 2000. The number of new psychoactive substances (NPS) reported to have been circulating in drug markets, for example, reached 755 in 2024, with 118 of these substances reported for the first time. 

The UNODC is not a libertarian organization. It is the UN agency charged with supporting and implementing the international drug control treaties and assisting member states in carrying them out. 

For years, I have argued that prohibition doesn’t eliminate drugs people want to use. It changes the incentives. As enforcement intensifies, producers and traffickers respond by developing more potent, more compact, and easier-to-conceal substances. The result is what prohibition critics have long called the “Iron Law of Prohibition”: the harder the enforcement, the harder the drugs.

Ironically, the latest UNODC report provides fresh evidence consistent with that argument. It describes traffickers exploiting new technologies, geopolitical instability, and an “unprecedented spike” in novel synthetic drugs to evade enforcement and expand into new markets. Far from disappearing under pressure, illicit drug markets continue to innovate. That innovation increasingly produces drugs that are more concentrated, more unpredictable, and more dangerous.

The report emphasizes that every major illicit drug market is becoming more global, more innovative, and more resilient, despite decades of increasingly aggressive enforcement. Methamphetamine production and trafficking have spread well beyond their traditional strongholds into the Middle East, Africa, Europe, and the Pacific. Cocaine production has quadrupled over the past decade as trafficking organizations expand into new markets in Africa and Asia. Even cannabis—long considered largely regional because it can be grown almost anywhere—is increasingly traded across continents, with North America emerging as a major source for international trafficking. 

The UNODC did not intend to validate my longstanding arguments against prohibition. Yet its latest report does exactly that. Governments can prohibit drugs, but they cannot prohibit demand. They can only change the incentives facing suppliers. Those incentives reward innovation, resilience, and increasingly potent, compact, and dangerous drugs. After decades of escalating enforcement, the evidence continues to point in the same direction: prohibition has not defeated the market. It has reshaped it.

When US residents barbecue and bask in the sun this July 4 to celebrate the 250th anniversary of “a new nation, conceived in liberty,” we still won’t be free to exercise one right that Europeans and others have been free to exercise for decades.

The US Food and Drug Administration (FDA) announced that it will no longer block manufacturers from selling the sunscreen ingredient bemotrizinol in the United States—yet the ban will remain in place until August 9. Why should you care that the FDA was banning sunscreens with bemotrizinol? Those products do a better job of blocking cancer-causing UVA rays than the sunscreens currently available to US consumers. They are more comfortable to use and thus offer still-greater protection because people will be more likely to use them. “This is overwhelmingly a positive development,” said Yale University professor of dermatology Mona Gohara. Sounds great, right?

Except, bemotrizinol has been available in Europe since 2000. The FDA has been blocking US consumers from using it all this time. While European authorities have been busy removing regulatory barriers to 30 different active sunscreen ingredients, the FDA had removed barriers to only 17—and none since 1999. Manufacturers have been petitioning the FDA to remove obstacles to bemotrizinol for over 20 years.

The FDA can’t claim that it has been keeping consumers safe. David Andrews, deputy director of the nonprofit Environmental Working Group, has remarked that relative to sunscreens currently available in the United States, “a number of these newer chemicals have a better safety profile in addition to better UVA protection.” If anything, the FDA has been undermining its own mission and making US residents less safe.

Cato Institute scholars have repeatedly criticized the FDA for denying US consumers their right to buy these products. George Mason University economist Alex Tabarrok has been banging on about FDA delay in approving new sunscreens since 2013. This insane situation has generated perplexed stories in major news outlets. And Rep. Alexandria Ocasio-Cortez (D‑NY; “AOC”) has complained that this a clear case of government overregulation.

Why did the FDA spend 20 years blocking this safe, effective sunscreen ingredient—and why does it still block so many others like it?

One reason is that the FDA is just responding to the incentives it faces. As I write in my latest book Recovery, 

Type I errors [read: blocking a harmful product] bring swift and certain retribution down on agency officials. The victims are easily identifiable.…Type II errors [read: blocking a beneficial product] bring almost no consequences for FDA officials. Even though delaying or blocking beneficial drugs can harm patients as much as approving unsafe drugs can.… Victims of Type II errors are much harder to identify. It appears the disease, not the FDA, killed them.

Due to this fundamental information asymmetry, the political system can discipline FDA officials only when their decisions cause patients to suffer or die from Type I errors. It effectively cannot discipline FDA officials when their decisions cause patients to suffer and die from Type II errors.

The FDA rationally focuses its energies on avoiding the harms that new products (might) pose and almost completely ignores the harms that the agency itself creates by blocking beneficial products. The result is that the FDA hurts more than it helps. I discuss one study of new drugs that found “the FDA was inflicting 12 times as much harm on patients through Type II errors as it was sparing patients by avoiding Type I errors.” My counterintuitive conclusion:

If FDA officials want to promote health, they should regulate less. They should approve new drugs faster and with less evidence of safety and effectiveness.

An interesting question is why, when it comes to sunscreens, the FDA appears to lean into these incentives more than European and other regulatory bodies that face the same incentives. A leading hypothesis is one that AOC mentions: Congress requires regulators to hold sunscreen ingredients to a higher regulatory standard (drugs) than European authorities do (cosmetics).

A second part of the answer is the way we talk about the FDA. Our everyday language hides what the FDA does and spares the agency the voter fury it has earned.

Start with a simple idea: You have a fundamental right to make your own health decisions. (Skeptics, see the Declaration of Independence’s second “among” and the Constitution’s Ninth Amendment.) When the FDA blocks you from buying sunscreens or other health products from willing sellers, it violates that right—along with others, including free speech. 

When the FDA removes those barriers, we call it an “approval.” But look at what actually happened: Government simply stopped blocking a product—stopped threatening to punish companies that sell it. We describe this as the FDA doing something helpful, when really it’s announcing that it will stop doing something harmful. When we say the FDA is “giving” consumers access to beneficial products, we credit government for removing harms the government itself inflicted. 

We also turn the Declaration on its head. We talk as if the freedom to make our own health decisions is something the government hands out—something regulators “allow.” All humans are created equal, but regulators who get to make other people’s health decisions for them are more equal than others.

Proceeding from a presumption of power (i.e., government should make these decisions) is no less an ideological choice than proceeding from a presumption of liberty (i.e., individuals should make these decisions). Only one of these presumptions appears in the Declaration and Constitution, though. These rhetorical conventions may simplify the prose, but they perpetuate the insanity.

————

So what should Congress do about it? In an ideal world, Congress would eliminate the FDA and other government interventions that prevent the health sector from solving the public goods problem the FDA does so much harm trying to address. 

Until that becomes politically feasible, Congress should remove barriers to US consumers purchasing health products available in other countries. Tabarrok writes, “Politicians as far apart as Sen. Ted Cruz, R‑Texas, and Rep. Alexandria Ocasio-Cortez, D‑NY, have indicated support for fast-tracking approval in the US for at least some drugs and devices already approved in other developed countries—a rare moment of wise bipartisan agreement.”

The danger is that the FDA will take the same sluggish approach to removing barriers to sunscreens that it has to removing prescription requirements. Medical authorities have begged the FDA to eliminate prescription requirements from, for example, all contraceptives and all forms of the opioid-overdose-reversal drug naloxone. Instead, the FDA has removed prescription requirements from just one oral contraceptive (the “mini pill,” in 2023) and two naloxone nasal inhalers (Narcan in 2023 and Rextovy earlier this month). Since President Trump pledged back in January that his “Great Health Care Plan” would make “more verified safe pharmaceutical drugs available for over-the-counter purchase,” the FDA has removed prescription requirements from only two medicines (Rextovy and the adapalene/​benzoyl peroxide acne gel Differin Epiduo).

Daily-use oral contraceptives have been on the market for 60 years. The American College of Obstetricians and Gynecologists and the American Academy of Family Physicians have been calling on the FDA to eliminate prescription requirements for all contraceptives for at least 12 years. The FDA’s deregulators need to pick up the pace.

Despite reports that they are less harmful than regular cigarettes, legislators have targeted e‑cigarettes and vapes via crackdowns and online sales bans, citing a desire to protect young people. 

A new study examines the effect of online sales bans on use of electronic nicotine delivery systems (ENDS). The study

used data from five national surveys conducted between 2013 and 2023 and leveraged the staggered adoption of the bans across states … [to] reveal no evidence that prohibiting online sales reduced youth e‑cigarette use. … Furthermore, these bans had a minimal effect on the frequency of use among continuing users.

The legislation was ineffective because

young people rarely used the internet to obtain ENDS products before the bans. … Although the bans significantly reduced online purchases of ENDS products, the overall reduction in youth use was less than 1 percentage point.

Indeed,

online purchases fell by 40–50 percent on average, suggesting that young people continued to obtain ENDS products online through illegal shipments … [and] shifted from online to in-person purchases and obtained more ENDS products from family and friends.

The bans also did not stop adults from using ENDS; the study shows 

no evidence that online sales bans reduced e‑cigarette or cigarette use among adults, even though adults use these products at higher rates than young people.

Cross-posted from Substack.

Earlier this week, the office of Representative Steve Cohen (D‑TN), who serves on the House Judiciary Committee, announced that he introduced HR 9186, “The Controlled Substances Act (CSA) Clarification in Sciences Act.” 

Under the CSA, the Drug Enforcement Administration (DEA) has established five schedules of controlled substances. The DEA defines Schedule I drugs as follows:

Schedule I drugs, substances, or chemicals are defined as drugs with no currently accepted medical use and a high potential for abuse. Some examples of Schedule I drugs are: heroin, lysergic acid diethylamide (LSD), marijuana (cannabis), 3,4‑methylenedioxymethamphetamine (ecstasy), methaqualone, and peyote.

Federal law prohibits the manufacture, sale, transportation, or personal possession of Schedule I drugs. Clinical researchers who want to study Schedule I drugs must go through a lengthy and discouraging approval process. Psychedelic drugs were placed on Schedule I in 1970, setting back promising research for half a century and driving most therapeutic research underground. Lawmakers and policymakers are only recently beginning to appreciate the therapeutic potential of psychedelics, which might have been helping patients decades ago.

The DEA considers drugs in Schedules II through V to have “accepted medical use(s).” The higher schedule numbers reflect decreasing “potential for abuse, with use potentially leading to severe psychological or physical dependence.”

The current scheduling regime relies on narrow definitions of “accepted medical use” and expansive concepts of abuse that can keep potentially beneficial substances under highly restrictive controls long after scientific understanding has evolved. Critics contend that the system has impeded research, restricted patient access, and allowed political considerations to influence scheduling decisions. H.R. 9186 attempts to correct these problems by updating key definitions, distinguishing dependence from abuse, recognizing a broader range of medical evidence, and giving greater weight to scientific and medical expertise in scheduling determinations.

Specifically, the bill recognizes that dependence is not synonymous with abuse, broadens the concept of accepted medical use beyond FDA approval alone, and gives greater weight to scientific and medical evaluations in scheduling decisions. It also raises the threshold for concluding that a substance lacks accepted safety, requiring regulators to consider both public health risks and legitimate medical uses when determining the appropriate level of control.

Rather than focusing primarily on the presence of risks, it requires regulators to weigh those risks against potential medical benefits and determine whether “a reasonable practitioner would conclude that the risks of death or significant irreversible physical or psychological harm to the user would clearly outweigh any medical benefit to the user.”

H.R. 9186 does not eliminate the underlying problem that the Controlled Substances Act places physicians and researchers within a system in which law-enforcement agencies retain substantial authority over medical decisions. In that respect, it still leaves cops practicing medicine. But by modernizing key definitions and placing greater emphasis on scientific and medical evidence, the bill would move federal drug policy in a more evidence-based direction. If enacted, these reforms could reduce barriers to research and improve clinicians’ ability to care for patients—including those in pain—while leaving broader questions about the role of law enforcement in drug policy unresolved.

AI in Health Care: A Policy Framework for Innovation, Liability, and Patient Autonomy—Part 9

Large language models are good at summarizing information. They are much less reliable at conveying scientific uncertainty.

Public health debates are increasingly being filtered through artificial intelligence. That becomes problematic when the science itself remains unsettled.

Consider the debate over 7‑hydroxymitragynine (7‑OH), a naturally occurring alkaloid in kratom. Ask several popular AI systems about the compound, and you’ll often get remarkably confident descriptions of its potency and risks. One AI-generated summary I recently encountered described 7‑OH as a “gas station opioid” that is many times more potent than morphine and poses serious overdose risks.

Readers encountering such claims could reasonably conclude that the science is settled. However, the scientific literature presents a more complicated picture.

The debate over 7‑OH has intensified as concentrated products have become more widely available. Kratom, a plant native to Southeast Asia, has been used for generations and is increasingly popular in the United States among people seeking pain relief or alternatives to more dangerous prescription opioids and illicit drugs. The controversy centers on concentrated 7‑OH products, which contain high levels of one of kratom’s active compounds. Critics argue that these products pose substantial risks and warrant stricter regulation. Others contend that public health messaging often overstates the evidence and relies on selective interpretations of the literature. Researchers continue to investigate the compound’s pharmacology, abuse potential, and safety profile.

What the Research Actually Shows

The question of potency alone illustrates the challenge.

The frequently repeated claim that 7‑OH is many times more potent than morphine often traces back to older animal and tissue studies. Yet more recent research on human μ‑opioid receptors has produced a more nuanced picture. In a 2021 study, researchers found that 7‑OH activates opioid receptors differently from conventional opioids and may not fit neatly into comparisons based solely on potency.

Other investigators have emphasized the distinct pharmacology of both 7‑OH and mitragynine, the primary active compound in kratom from which 7‑OH is derived, including signaling pathways that differ from those of many conventional opioids. The work of Kruegel and colleagues helped establish that mitragynine and 7‑OH exhibit atypical opioid receptor signaling, raising questions about simple potency comparisons with conventional opioids.

The discussion of overdose risk is similarly more complicated than many AI-generated summaries suggest. Concentrated 7‑OH products raise legitimate safety concerns, yet fatal overdoses involving 7‑OH appear to be uncommon. Reported deaths often involve multiple substances, leaving researchers to debate what role, if any, 7‑OH played.

The existence of these competing interpretations does not establish that concentrated 7‑OH products are harmless. Legitimate concerns remain about dependence, misuse, product quality, marketing practices, and long-term health effects. Researchers are still working through many of these questions.

The scientific literature reflects that ongoing debate. AI-generated summaries often do not.

Why AI Struggles With Scientific Uncertainty

Large language models are trained to synthesize information from vast amounts of text. They excel at condensing complex subjects into concise explanations. They do not reliably distinguish between settled scientific questions and ongoing scientific controversies. When researchers disagree, AI systems frequently default to the interpretation most prevalent in their training data and may understate the extent of scientific disagreement.

In debates over emerging substances, that dynamic can accelerate moral panics. Repeated claims about potency, overdose risk, or abuse potential may acquire an aura of certainty long before the underlying science has reached a consensus.

That tendency can be especially misleading in public health, where evidence evolves, and experts often disagree on how risks should be interpreted. Questions about nutrition, nicotine alternatives, psychedelics, environmental exposures, and emerging therapies often generate competing interpretations of the same evidence.

Scientific progress emerges through debate, reanalysis, and the testing of competing explanations. Researchers challenge assumptions and revise their conclusions. By contrast, AI systems are optimized to provide answers.

That tendency creates a risk that uncertain claims appear certain simply because they are repeated frequently enough in the underlying information ecosystem.

The Wrong Solution: Regulating Scientific Disagreement

Some observers have responded to concerns about AI misinformation by calling for greater regulatory oversight of AI outputs, trusted information frameworks, and government involvement in defining responsible AI behavior. However, public health institutions have their own histories of error, overconfidence, and political influence. Empowering regulators to referee scientific controversies risks replacing one source of false certainty with another.

Artificial intelligence holds tremendous promise for health care and scientific research. These tools can help clinicians navigate the medical literature, assist researchers in identifying patterns, and help the public access information that would otherwise remain buried in journals and databases.

Yet many public health controversies already involve institutions presenting uncertain evidence as settled fact. AI systems trained on that information ecosystem may inherit the tendency to understate uncertainty and project confidence beyond what the evidence warrants.

Competition among models, transparency about uncertainty, open debate, and continued scrutiny of confident claims offer a healthier path forward. That scrutiny should apply whether those claims originate with advocacy groups, regulators, journalists, or AI systems.

The debate over 7‑OH will continue as new evidence emerges. Public health discussions are healthier when uncertainty remains visible. Policymakers considering restrictions or bans should remember that today’s controversy may become tomorrow’s accepted therapy. Many substances now used in medicine were once dismissed, feared, or poorly understood. 

AI can summarize a scientific controversy, but it cannot resolve one. When the evidence remains contested, confidence is not the same thing as certainty—and a chatbot’s answer is not a substitute for scientific debate.

To read other parts of this blog series, go here.