The End of NRC Independence and the Future of Nuclear Safety

One month ago, the U.S. Supreme Court in Trump v. Slaughter overturned Humphrey’s Executor v. United States, stripping the U.S. Nuclear Regulatory Commission (NRC) and agencies like it of the structural independence that they have relied on for decades. The Commission is now a “normal” executive agency subject to direct presidential control—full stop, no exceptions for nuclear safety. The implications have yet to set in for the already-overwhelmed NRC, the industry it regulates, or the public. In this eerie interval between the hit and the reckoning, it is worth taking stock: What has happened, what it means for coming years, and what might still be done to reduce the risk of nuclear disaster.

The blow fell on June 29, 2026, when the Court sustained President Donald J. Trump’s firing of Federal Trade Commission Commissioner Rebecca Slaughter. Slaughter challenged her firing as unlawful: The Federal Trade Commission Act provides that Commissioners can only be removed by the President for “inefficiency, neglect of duty, or malfeasance in office.” The Supreme Court had upheld this very provision of the Act in 1935, in Humphrey’s, against a separation-of-powers challenge.

The Court in Humphrey’s gave credit to the U.S. Congress’s judgment that, to do the work assigned to them by statute, some agencies benefit from a distinctive structure–specifically, a multimember board of principal officers who could be removed (that is, fired) by the President only “for cause,” not merely for political or personal reasons. And these so-called independent agencies—the Federal Trade Commission, the Federal Communications Commission, the Securities and Exchange Commission, the Consumer Products Safety Commission, and dozens more—were a fixture of American government for nine decades.

Humphrey’s, typically one of the first cases taught in an administrative law course, was foundational for modern administrative law. An entire generation was born and died under the constitutional structure that stabilized under this system. Over that time, Congress relied repeatedly upon the equilibrium the Court had created in Humphrey’s to stand up new independent agencies when it did not feel it was wise to delegate more directly to the President.

Enter the Roberts Court. Humphrey’s, the Court said by a vote of 6-3, is and always was wrong: Congress does not and never did have the power to insulate any agency from presidential whim (except, apparently, for the Board of Governors of the Federal Reserve System, the Supreme Court added in the companion case Trump v. Cook). On the majority’s telling, Slaughter is not a revolution but a restoration: a move to return Article II of the U.S. Constitution to its proper shape by reuniting executive power with executive control. To the majority, this is an exercise in housekeeping rather than demolition, sweeping away decades of doctrinal clutter to uncover the Constitution’s original design.

The independent NRC is one of many agencies to be Slaughter-ed. The constitutional law backdrop Congress legislated against when it set up the agency in 1974 is now gone, as it is for every agency that lacks the Federal Reserve’s American founding-era pedigree.

For the NRC specifically, presidential control is not an abstract idea. Even before the case, the President had taken multiple actions to exert control over the agency, including issuing multiple executive orders targeting it and firing a sitting Commissioner, Christopher Hanson, without waiting for the Supreme Court’s blessing. Any commissioner who declines to fall in line with the President’s agenda can expect the same treatment.

So, what now?

Unless other actors rise to this new challenge – exhibiting new resolve – the probability of nuclear accidents is now greater than it was before Slaughter.

Independence is not the sole determinant of nuclear safety, and doing away with independence does not make a nuclear “incident” inevitable. It does, however, put greater pressure on other safeguards. And where those other safeguards are already pressured or frayed, a loss of independence may tip the balance. The post-mortem explanation for the future disaster writes itself: A safety regulator under extraordinary strain, overseeing an industry undergoing rapid technological change, faces increased political pressure to bring reactors online and then loses the capacity to resist that pressure. Boom. To keep the probability of nuclear accidents from rising after the loss of agency independence, we would need to see compensatory moves from other actors and institutions involved in nuclear safety.

To understand the importance of the loss of independence, it is important to understand the status of nuclear safety even prior to Slaughter. Specifically, the NRC was already an agency in overwhelm—struggling to stay afloat while addressing extraordinary transformations in technology, particularly in advanced and small modular reactors, while trying to make sense of new legal authority, namely, the ADVANCE Act, which changed the agency’s mission, and Loper Bright Enterprises v. Raimondo, which reduced the agency’s authority to interpret its own statutory mandates.

Now this overwhelmed safety regulator has been stripped of its structural independence and will be pressured by a newly empowered President who has already made clear his priorities for faster licensing. The path of least resistance for the agency is to start rubber-stamping licenses for a complex technological transition to advanced and microreactors. And the nuclear industry submitting these licenses is long conditioned to rely on the formerly precautionary agency to ensure that its activities were adequately safe. Many players in the industry have lost both the norms and the training to perform their own substantive risk analyses.

The result is, admittedly, not inevitable, but the post-mortem explanations for a future disaster write themselves: a safety regulator under extraordinary strain overseeing an industry undergoing rapid technological change with fewer institutional guardrails and, now, less capacity for independent judgment.

How can we choose a different future? The critical first move is to be clear-eyed about the situation: The NRC is no longer independent. In light of some remaining confusion in the industry, it is worth considering a few counterarguments that may occur to some readers and explain why they are wrong.

One might try to hang some hope on Cook: The Federal Reserve is still independent, so perhaps the NRC is too. But that is an empty hope. The kind of history the Supreme Court wants as evidence of agency independence is founding-era history, like the Supreme Court perceives in central banks. You know: founding history—the cool history (although not everyone agrees that the founding-era history supports a central bank exception).

This is bad news for the NRC, at least if one likes agency independence: It has only been 51 years since Congress, facing what was widely viewed as the failure of prior nuclear regulatory regimes under the Atomic Energy Commission, purposefully bifurcated authority to regulate nuclear safety. Congress distinguished between what was sensible to leave to the President—including promotion of nuclear power, now delegated to the cabinet-level U.S. Department of Energy—and what had historically gone poorly when combined with industry pressure and politics: setting nuclear standards and granting licenses, which Congress delegated to the newly created independent NRC. If only Alexander Hamilton had commented on the importance of independent nuclear safety, perhaps the agency could have had its own Cook-style precedent, and we would not be in such a pickle now. But if the FTC is too young for independence at 111 years, how much further from the age of majority is the NRC, at a youthful 52?

Nor can the NRC claim independence on the grounds that its work is too important and technical to trust to political or industry-pressured processes, though those these were justifications for the agency’s creation by Congress in 1974. Indeed, the court in Slaughter is clear that, except for the Federal reserve, only agencies lacking real power may be independent. As the court explains, “all that is left of Humphrey’s is its observation that an agency that ‘exercises no part of the executive power’ need not fall within the rule of Presidential removal.”

No, for formerly independent agencies so unwise as to lack founding-era history, the cat is among the pigeons now. And surprise: At least some of the pigeons are radioactive.

So what is left?

Justice Neil Gorsuch, alone among the Slaughter majority, appears to recognize the mess that the Court has created, writing that if “the task of fixing a problem belongs to those who made it, this Court has some work to do.” Congress loaded independent agencies up with power on the Court’s assurance that that the President could not direct how that power was used, but now he can. Oops.

But, Justice Gorsuch reassures us, the judiciary has “no shortage of tools” to help; the courts can lean on the non-delegation doctrine, the major questions doctrine, and Loper Bright, to keep the executive in its constitutional place. In other words, his prescription is to continue taking apart the administrative state.

Can the courts get us nuclear safety, either with these tools or others? Our optimism should be constrained. It is true that courts remain our most powerful check on executive overreach through ensuring that agencies comply with the law. For example, suppose the President directs the NRC to license advanced reactors so quickly that it is unsafe. After unsafe licenses are granted and challenged, courts will be there to review and require fidelity to the NRC’s statutory obligations—to ensure “reasonable assurance of adequate protection to public health and safety” while regulating in a manner that is “efficient.” Justice Gorsuch is right that the courts’ authority in such cases is only strengthened by Loper Bright, which blocks courts from going along with an agency’s—or perhaps more importantly a President’s—“close enough” interpretation of what a statute requires.

That said, courts cannot do it all. Courts must review agency statutory interpretations independently after Loper Bright, but not all aspects of a licensing decision get that level of scrutiny. Even if they did, it is unclear that an inexpert, generalist court will do better than an untrained President at making complex safety tradeoffs. Not to mention that any checks courts provide are likely to be retrospective and will necessarily involve some guesswork on the part of current actors, both because the Supreme Court will likely continue to break new legal ground and because the new statutory mandates of the agency have yet to be interpreted by any court. This is a difficult and incomplete solution set for an industry actively grappling with transformative and potentially catastrophic new technologies.

What, or who, is a solution then? There are at least three actors who might be able to help reduce the chance and magnitude of pending disasters.

First is the NRC itself. The key here is for the agency to accept its new legal position—and then place that acceptance in context. The Commission is now an executive agency with no insulation from presidential removal, yes—but it is an executive agency that still has statutory responsibilities constitutionally delegated by Congress. The key is to take those responsibilities seriously and not fall into the trap of anticipatory obedience, placing the President’s agenda above fealty to its legislative obligations. The NRC’s statutes—including the Atomic Energy Act of 1954, the Energy Reorganization Act of 1974, and the ADVANCE Act of 2024—should be its backbone. Whatever else the President directs, the NRC remains responsible for assuring “reasonable assurance of adequate protection to public health and safety,” while regulating in a manner that is “efficient,” considering benefits to society as well as costs. Loper Bright should serve as a shield against at least the extremes of executive control. Whereas under Chevron, Presidents could have directed many possible statutory interpretations depending upon their preference, both the President and the agency are now bound to the “best” interpretation as determined by courts. The NRC must now play a long game that is familiar to many other executive agencies: balancing the immediacy of satisfying the current President with both the potential of judicial review and the knowledge that a new President will come in time. Whatever policies it adopts must withstand that change as well as the present one.

Second, the nuclear industry will need to grow up fast. For many decades, the precautionary posture of the NRC—including its hesitancy to approve new technologies or methods—led the industry towards a culture of dependency. Now, like children freed from the strictures of an attentive nanny, the temptation for industry will now be to run wild, luxuriating in new-found freedom from the oppression of someone else’s planning, risk analysis, and safety-based procedures. This freedom is emerging at the same time as there is extraordinary forward momentum—and money to be made—in licensing new advanced reactor technologies, many of them the first of their kind. The industry can no longer rely upon the NRC to make sure that these technologies are “safe enough.” It must learn to take responsibility for itself, investing in probabilistic safety methods and reflective risk analysis. Failure to take those steps now will result in more, and worse, accidents down the line—accidents that are not only bad for individual industry actors but that harm the industry and the public as a whole.

In addressing industry risk, it is worth noting that insurers, particularly the American Nuclear Insurers, may have a particularly critical role to play. The Price-Anderson Nuclear Industries Indemnity Act of 1957 requires nuclear operators to maintain private insurance and participate in pooling. Insurers should, and we can expect that they will, respond to Slaughter with substantial increase in premiums—especially where industry actors are unable to show that they have stepped up to address risk analysis carefully, even absent regulatory backstops.

Finally, Congress has a potential role to play for the NRC, as with other formerly independent agencies. Congress may want to review statutory provisions that the NRC administers with an eye to removing any ambiguity about what it expects from the agency. Congress’s oversight role also rises in importance as its power to insulate falls. The Subcommittee on Clean Air, Climate, and Nuclear Innovation and Safety of the Senate Committee on Environment and Public Works, and the House Committee on Energy and Commerce’s Subcommittee on Energy share oversight responsibilities for the NRC. Congressional hearings are no magic bullet, but they can help to spotlight how the NRC is carrying out its responsibilities.

When Congress established the NRC as an independent agency in 1974, it legislated against a constitutional law backdrop that has now vanished. To outward appearances, the agency may look much like it did three weeks ago, but it is now a different institution. Between the agency’s existing capacity issues, the presidential pressure to bring reactors online, and the fact that the industry has long outsourced responsibility for safety to its watchful regulator, the prospect of a nuclear accident is more foreseeable today—not because nuclear technology has become inherently more dangerous but because the institutional assumptions on which Congress relied when in created an independent safety regulator have materially changed. Humphrey’s is not coming back, but there are still steps that the agency, the industry, insurers and Congress can take to help push risks back down.

Arden Rowell is a professor of law at the University of Illinois Urbana-Champaign College of Law.

Jud Mathews is a professor of law and political science at Penn State Dickinson Law.