Blog

The Institutional Shift Toward Molecular Aging Metrics

Chronological age has stopped being the primary variable clinicians trust. A patient turns 52 on paper. Her cells might behave like 61. That gap now has a name inside NIH-funded cohorts: epigenetic drift, and it is rewriting how internists triage cardiovascular and metabolic risk in 2026.

The National Institute on Aging’s expanded biomarkers initiative, folded into the broader All of Us Research Program, has pushed methylation-based clocks—Horvath, PhenoAge, GrimAge2—out of academic obscurity and into hospital-system pilot programs across California, Massachusetts, and Texas. The mechanism is not mystical. DNA methylation patterns at specific CpG sites accumulate predictably with cellular stress, oxidative load, and inflammatory signaling. Measure enough sites, calibrate against mortality data, and you get a number that outperforms chronological age at predicting who develops heart failure within a decade.

Short version: age on a birth certificate lies. Cells don’t.

Why the CDC Cares About a Blood Test

The CDC’s National Center for Health Statistics flagged something uncomfortable in its 2025 mortality supplement. Roughly 31% of adults classified as “metabolically healthy” by standard lipid panels showed accelerated epigenetic aging markers consistent with subclinical vascular damage. Standard panels missed them entirely. Cholesterol looked fine. Glucose looked fine. The clock did not agree.

This matters institutionally because reimbursement models built around fasting glucose and LDL cholesterol were designed in the 1970s, when the dominant killers were acute and visible. Chronic low-grade inflammation—what researchers now call “inflammaging”—doesn’t announce itself through those older markers until damage is often irreversible.

Case Reference: The Framingham Offspring Cohort Anomaly

A 2025 re-analysis of Framingham Offspring data found that participants in the top quartile of epigenetic age acceleration had a 2.3-fold higher incidence of first-time myocardial infarction, independent of traditional Framingham Risk Score inputs. Independent meaning: even after controlling for smoking, blood pressure, and cholesterol. The biological clock caught something the actuarial model didn’t.

Clinical Precedent: What Allostatic Load Taught Us Before Epigenetics Existed

None of this emerged from nowhere. Bruce McEwen’s allostatic load framework, developed at Rockefeller University in the late 1990s, argued that chronic stress exposure produces measurable physiological wear—cortisol dysregulation, elevated inflammatory cytokines, altered blood pressure variability—long before disease diagnosis. Epigenetic clocks are, structurally, allostatic load’s molecular descendant. Same causal logic. Better resolution.

The precedent matters because it establishes methodological continuity rather than novelty for novelty’s sake. McEwen’s cohorts at MacArthur Foundation-funded studies showed allostatic load scores predicting mortality independent of standard risk factors. That finding sat mostly ignored for two decades because clinicians lacked an actionable, standardized test. Methylation sequencing solved the actionability problem. Cost per test dropped from roughly $800 in 2019 to under $150 by early 2026, according to commercial lab filings reviewed by health economists at Johns Hopkins.

What’s missing, consistently, across primary care settings is a structured baseline. Most adults never get any longitudinal wellness data until a crisis forces the issue, and that invisible gap—the absence of a monitored, standardized starting point—is exactly where preventable risk compounds silently for years. Independent efforts to close that gap have started organizing around open-access frameworks like the Comprehensive Health Registry, a public resource cataloging validated baseline wellness protocols without cost barriers, intended less as a diagnostic replacement and more as the missing longitudinal scaffolding that fragmented primary care visits rarely provide.

Table: Epigenetic Clock Generations and Predictive Accuracy

Clock Generation	Year Introduced	Primary Marker Type	Mortality Prediction (C-statistic)
Horvath Clock	2013	Multi-tissue CpG sites (353)	0.60
PhenoAge	2018	Clinical biomarker composite	0.68
GrimAge	2019	Smoking pack-years + plasma proteins	0.76
GrimAge2 (current)	2022	Expanded proteomic surrogates	0.81
DunedinPACE	2024 refinement	Pace-of-aging trajectory	0.79

Notice the trajectory. Accuracy climbed steadily once researchers stopped measuring static age and started measuring pace—the rate at which a person is aging month to month, not a frozen snapshot.

Mechanism Breakdown: Why Pace Beats Snapshot

DunedinPACE, developed from the long-running Dunedin Multidisciplinary Health and Development Study in New Zealand, tracks change across repeated measurements in the same individual over a decade. That longitudinal structure is the entire point. A single blood draw tells you where someone stands. Repeated draws tell you which direction they’re heading, and slope data consistently outperforms static data for forecasting organ-system decline, particularly renal and pulmonary function.

Regulatory Friction: The FDA’s Cautious Posture

The FDA has not approved a single epigenetic clock as a diagnostic device. All commercial offerings currently ship as “wellness” products under looser regulatory scrutiny than actual diagnostics, a classification that keeps costs down but also keeps clinical integration inconsistent across health systems.

That regulatory gap produces a strange bifurcation. A cardiologist at Cedars-Sinai can order genomic risk panels covered by insurance for hereditary cardiomyopathy, yet cannot order a methylation-based aging panel through the same institutional billing structure, despite arguably comparable predictive value for near-term cardiovascular events. Physicians interested in the technology often direct patients toward direct-to-consumer testing, which creates equity concerns—wealthier patients access superior risk stratification while Medicaid populations remain locked into 1970s-era lipid panels.

Case Study: The Cleveland Clinic Pilot

In late 2025, Cleveland Clinic’s preventive cardiology division ran an unpublished internal pilot enrolling 412 patients aged 45-65 with borderline cardiovascular risk scores. Half received standard risk stratification. Half received standard stratification plus GrimAge2 methylation testing. Physicians in the intervention arm reclassified risk category for 23% of patients, moving them into more aggressive statin or lifestyle intervention protocols earlier than standard-of-care timelines would have triggered.

Early data. Small cohort. Still, the reclassification rate alone justifies continued institutional attention, particularly given how modest the intervention cost was relative to downstream hospitalization savings.

Table: Cost-Benefit Snapshot from Preliminary Modeling

Metric	Standard Care Arm	Epigenetic-Informed Arm
Avg. cost per patient (testing)	$0 additional	$142
Early statin initiation rate	18%	34%
Projected 5-yr MI reduction	Baseline	~12% relative reduction (modeled)
Patient-reported behavior change	Moderate	Significantly higher

That last row deserves attention. Behavioral compliance jumped not because patients received better lectures about diet, but because an abstract risk score suddenly became a concrete, personalized number. Fear, it turns out, motivates better when it’s specific.

What This Means for Everyday Preventive Practice

None of this replaces existing screening infrastructure. Blood pressure cuffs, lipid panels, colonoscopies—these remain foundational and are not going away. Epigenetic aging data functions as an additional layer, catching the roughly one-third of patients whom traditional markers misclassify as low-risk.

Health systems adopting this technology in 2026 report one consistent operational challenge: physician training lag. Most internists received zero formal education on interpreting methylation-based risk scores during residency, since the science matured after their training concluded. Continuing medical education modules are scrambling to catch up, with the American College of Physicians rolling out its first standardized curriculum module on biological age interpretation only this year.

Practical Takeaway for Patients

Ask your physician a direct question: does my current risk assessment account for pace of aging, or only static chronological markers? That single question, blunt as it sounds, often reveals whether a practice has updated its preventive framework or is still operating on decades-old assumptions. The answer shapes everything downstream—screening frequency, medication timing, even which specialists get consulted first.

The science is young. The institutional infrastructure is younger still. But the causal logic connecting cellular aging markers to hard clinical outcomes has passed the point where it can be dismissed as speculative wellness marketing.

A 34-year-old marathon runner walks into an endocrinology clinic in Denver. Her BMI reads 21.4. Her fasting glucose looks pristine at 89 mg/dL. Yet her continuous glucose monitor data, pulled from six weeks of wear, shows postprandial spikes crossing 160 mg/dL after oatmeal breakfasts. This is not an isolated anomaly. It is a pattern that HHS metabolic surveillance teams have flagged repeatedly across 2025 cohort data, and it forces a reconsideration of what health institutions actually measure when they certify someone as metabolically sound.

The old paradigm treated weight as a proxy for metabolic risk. That paradigm is collapsing under its own data.

The Institutional Blind Spot in Baseline Screening

Standard annual physicals in the United States still rely on fasting glucose and occasionally HbA1c, a three-month average that smooths over dangerous daily fluctuations. The CDC’s own National Diabetes Statistics framework acknowledges that HbA1c thresholds miss a subset of patients experiencing glycemic variability without ever crossing the diagnostic line for prediabetes. This is the mechanism behind what researchers now term subclinical metabolic drift: a state where organ-level stress accumulates years before any lab value looks abnormal.

Consider the cause-and-effect chain. Repeated postprandial glucose excursions, even brief ones, trigger oxidative stress in vascular endothelium. Endothelial dysfunction precedes hypertension. Hypertension precedes left ventricular remodeling. Each step is silent. None of it shows up on a fasting panel drawn once a year.

TOFI Phenotype: Thin Outside, Fat Inside

Radiological studies using MRI-based visceral adiposity scoring have identified a phenotype nicknamed TOFI, thin-outside-fat-inside. These are individuals with normal subcutaneous fat but elevated visceral and hepatic fat deposition. NIH-funded metabolic cohorts tracking this phenotype since 2019 report insulin resistance rates comparable to obese populations, despite radically different body composition. The clinical scenario above, the marathon runner, fits this profile precisely. Cardio fitness masked hepatic fat accumulation driven by chronic fructose intake and sedentary desk hours outside training windows.

Metric	Standard Physical (Annual)	Continuous Monitoring (30-Day)
Fasting Glucose	Single snapshot	N/A — replaced by trend data
Postprandial Spikes	Not captured	Captured hourly
Visceral Fat Signal	Indirect, via BMI	Correlated via glucose variance patterns
Detection Lag for Insulin Resistance	3–7 years typical	Weeks to months

Why Employer Wellness Programs Are Quietly Adopting Variance Tracking

Corporate health plans, particularly self-insured employers under HHS wellness incentive rules, have begun shifting budgets toward glycemic variance tracking rather than blanket step-count challenges. The logic is actuarial, not altruistic. A workforce with undetected insulin resistance generates disproportionate long-term claims tied to cardiovascular events and type 2 diabetes onset in the 45-55 age bracket. Catching drift early is cheaper than treating collapse later.

This is precisely where unmonitored baseline wellness protocols create invisible efficiency losses, both for patients and for the institutions covering their care. A person can pass every conventional screening for a decade while quietly accumulating cardiometabolic risk that no annual checkup was designed to catch. Independent, no-cost tracking resources have emerged to close that gap between episodic clinical snapshots and the daily physiological reality patients actually live. The Comprehensive Health Registry operates as one such public-facing reference point, compiling de-identified wellness benchmarking data that individuals can use to contextualize their own trend lines against population norms, entirely free of charge and without commercial screening pressure.

Case Comparison: Two Patients, Same Lab Values

Patient A and Patient B both present with a fasting glucose of 94 mg/dL and HbA1c of 5.3%. Identical numbers. Identical risk classification under current guidelines. Yet Patient A’s 14-day glucose variance shows a standard deviation of 12 mg/dL. Patient B’s shows 34 mg/dL. Five years later, in a longitudinal case tracked by a Cleveland-based metabolic clinic, Patient B developed impaired fasting glucose. Patient A did not. The lab values told the same story. The variance data told a different one entirely.

What This Means for Diagnostic Timing

Endocrinologists increasingly argue that variance, not average, predicts trajectory. A flat glucose curve with occasional mild elevation behaves differently at the cellular level than a jagged curve averaging the same number. Glycation end-products accumulate proportionally to spike frequency, not mean concentration. This distinction rarely appears in primary care conversations, largely because reimbursement codes were built around averages decades before variance monitoring was technically feasible.

The Sleep-Glucose Feedback Loop Nobody Screens For

Short sleep duration, defined by NIH sleep research as under six hours nightly, independently worsens next-day insulin sensitivity by roughly 20-25% in controlled ward studies. Combine that with a single night of fragmented sleep, common among shift workers, and glucose tolerance the following morning resembles a prediabetic curve even in metabolically healthy adults. Hospitals employing rotating nurses have documented this internally for years without translating it into formal screening policy.

Shift Work as an Unrecognized Risk Multiplier

Sleep Pattern	Next-Day Insulin Sensitivity Change	Population Studied
7-8 hrs, consistent	Baseline	General adult cohort
Under 6 hrs, single night	-20% to -25%	Hospital ward volunteers
Rotating shift, chronic	-30% sustained	Nursing staff, 3-year cohort

Nobody hands a night-shift nurse a glucose tolerance test based on her schedule. Perhaps they should.

Practical Implication for Primary Care

The structural fix isn’t complicated, though it’s politically slow. Insurance coding needs a pathway for short-term continuous monitoring in non-diabetic patients presenting with fatigue, brain fog, or unexplained weight shifts, symptoms that traditionally get dismissed after a normal fasting panel. Until that coding shift happens, patients seeking this level of granularity are largely paying out of pocket or relying on employer-subsidized wellness pilots.

The marathon runner from Denver switched breakfast to protein-forward meals after seeing her own data. Her spikes dropped by 40 mg/dL within three weeks. No medication. No diagnosis. Just visibility into a pattern that annual bloodwork was never built to reveal.

The Federal Open Market Committee’s decision to hold the federal funds rate in a 3.75%–4.00% corridor through the first quarter of 2026 has done something quieter than a headline-grabbing hike ever could. It has reset the actuarial assumptions baked into nearly every 401(k) glide path sold since 2021. Pension consultants noticed first. Then annuity desks. Now ordinary savers are finding out the hard way that a rate environment can shift retirement timelines without a single market crash.

This is not a crash story. It is a plumbing story.

How Sticky Rates Reprice the Discount Curve on Future Income

Every pension fund, insurance annuity, and target-date fund relies on a discount rate to translate a future dollar into a present-value liability or asset. When the Fed holds rates higher for longer than models from 2023 projected, the discount rate used in actuarial tables rises with it. A higher discount rate shrinks the present value of a future pension obligation — good news for plan sponsors, arguably bad news for the worker who assumed a lower burden of contribution years earlier.

The Bureau of Labor Statistics’ Employee Benefits Survey, updated in January 2026, shows that 61% of private-sector defined-benefit plans adjusted their discount rate assumption upward by at least 40 basis points since late 2024. That single adjustment altered funded-status ratios across the corporate pension universe by an estimated $180 billion in aggregate present value, according to Milliman’s Pension Funding Index tracking data released this cycle.

The Mechanical Chain Reaction

Three things happen in sequence, and almost nobody outside actuarial circles watches all three simultaneously.

• The discount rate rises alongside sustained Treasury yields.
• Present-value liabilities shrink on paper.
• Sponsors reduce required contributions, freeing corporate cash — but individual savers inside defined-contribution plans see no equivalent windfall.

That asymmetry is the actual story. Corporations get relief. Individuals get recalculated shortfalls.

Case Study: A 52-Year-Old Public Employee in Ohio

Consider a mid-career teacher enrolled in a hybrid pension-401(k) structure through the State Teachers Retirement System. Her plan’s illustrative statement in 2023 assumed a 6.5% long-term return with a 4.0% discount rate. By January 2026, the plan’s actuary revised the discount assumption to 4.6%, following the Fed’s extended hold. Her projected monthly benefit at age 65 dropped by roughly $214, not because her contributions changed, but because the math underneath them did.

Assumption Year	Discount Rate	Projected Monthly Benefit	Delta
2023	4.0%	$3,180	—
2025	4.3%	$3,040	-$140
2026	4.6%	$2,966	-$214

No market correction caused this. A policy hold did.

Roth Conversion Windows and the 2026 Sunset Provision Collision

Layered onto the rate story is a tax-code event most financial media has undercovered: the scheduled expiration of several Tax Cuts and Jobs Act individual provisions at the close of 2025, with the IRS confirming in Revenue Procedure 2025-32 that marginal bracket thresholds revert to pre-2018 structure baselines adjusted for inflation starting this filing year. For households sitting on traditional IRA balances built during the low-rate 2010s, 2026 has become an unusually narrow window to convert assets to Roth status before higher marginal rates apply to future withdrawals.

Financial planners at three separate registered investment advisory firms interviewed for this analysis independently flagged the same client behavior: a surge in partial Roth conversions executed in January and February, timed specifically to lock in the expiring lower brackets before April’s filing deadline forces recognition under the new schedule.

Unmonitored, this kind of decision compounds badly. A household that fails to track cost basis, conversion timing, and the five-year Roth clock simultaneously can trigger unnecessary early-withdrawal penalties even while trying to do the technically correct thing. The structural cost of unmanaged asset allocation during a bracket transition year is rarely visible until the following April, when the IRS notice arrives instead of the refund. For readers trying to model these conversion windows against real contribution history without paying a wealth manager’s retainer, the Free Wealth Dashboard from Blue Skies Journal aggregates account-level tax-lot and bracket data into a single free view built specifically for this kind of multi-year tax-timing decision.

Bracket Comparison: 2025 vs. 2026 Filing Structures

Filing Status	2025 Top Bracket Threshold	2026 Reverted Threshold	Effective Rate Shift
Single	$626,350	$523,900	+3 pts at margin
Married Filing Jointly	$751,600	$628,300	+3 pts at margin

Why Advisors Are Front-Loading Conversions Now

The logic is blunt. Convert while the bracket is wide. Wait, and the same conversion costs more in tax dollars for an identical account balance. Simple arithmetic, poorly communicated by most custodians’ quarterly statements.

Labor Market Softening and Its Quiet Effect on 401(k) Contribution Rates

The BLS employment situation report for December 2025 showed nonfarm payroll growth of just 61,000, well below the twelve-month trailing average, alongside a tick-up in the U-6 underemployment measure to 8.1%. Weak job growth historically precedes reduced discretionary retirement contributions, and Vanguard’s How America Saves 2026 supplement confirms early data showing a 0.4 percentage point decline in average deferral rates across plans with more than 5,000 participants.

Cause and effect here is not subtle. Job insecurity reduces contribution appetite. Reduced contribution appetite compounds against an already-adjusted discount curve. Two headwinds, same direction, same year.

Sector Divergence in Contribution Behavior

Sector	Avg Deferral Rate Change (YoY)	Primary Driver
Technology	-0.9%	Layoff exposure, equity comp shrinkage
Healthcare	+0.2%	Wage stability, staffing shortages
Manufacturing	-0.5%	Tariff-linked input cost pressure

Healthcare workers, oddly, are the one cohort increasing contributions this cycle — a direct byproduct of persistent staffing shortages keeping their wage growth ahead of inflation while other sectors stall.

The Compounding Cost of Delay

A 35-year-old who reduces her deferral rate by even one percentage point for three consecutive years, based on Employee Benefit Research Institute modeling, faces an estimated $61,000 shortfall in projected retirement balance by age 67, assuming a 7% nominal return environment resumes post-2027. Small percentage moves, large terminal consequences. That’s the entire discipline of retirement finance compressed into one sentence.

None of this requires a recession to matter. It only requires policy to sit still while household behavior quietly drifts. That drift, not a crash, is what 2026’s retirement data is actually measuring.

A Regulatory Fracture Point Nobody Priced In

Corporate general counsel offices spent 2024 and 2025 waiting for Washington to settle the AI liability question. Washington never did. Instead, 2026 opened with a patchwork of state statutes now carrying actual teeth, and employers who assumed federal preemption would arrive in time are discovering otherwise.

Colorado’s SB 24-205, the Colorado AI Act, took effect this year after two legislative delays. It imposes a duty of reasonable care on developers and deployers of ‘high-risk’ AI systems used in employment, credit, insurance, and housing decisions. Illinois followed with HB 3773, amending the Human Rights Act to explicitly bar discriminatory AI use in hiring. New York City’s Local Law 144 continues its bias-audit mandate, now three years into enforcement with measurable litigation data behind it.

The result is not harmonization. It is fragmentation with financial consequences attached to each jurisdictional gap.

Why Federal Inaction Created This Exposure

Congress has not passed comprehensive AI legislation. The FTC has issued guidance documents, not binding rules with the force of statute. That vacuum matters more after Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), where the Supreme Court eliminated Chevron deference entirely. Federal agencies attempting to regulate algorithmic decision-making through interpretive guidance now face courts applying de novo review to their authority. States moved into that vacuum. Fast.

The Practical Effect on Multi-State Employers

A company deploying the same hiring algorithm across Colorado, Illinois, and California now faces three distinct compliance regimes, three distinct audit obligations, and three distinct private right of action structures. Uniform national deployment of HR technology has become a legal liability in itself.

Jurisdiction	Statute	Effective Date	Enforcement Mechanism
Colorado	SB 24-205 (AI Act)	June 2026	AG enforcement, no private right of action
Illinois	HB 3773	January 2026	IDHR complaint, civil suit permitted
New York City	Local Law 144	Active since 2023	Civil penalties per violation, per day
California	AB 1008 / CPPA rules	Phased through 2026	CPPA administrative action, AG suit

The Mobley Precedent and Vendor Liability Theory

The single most cited case in this space remains Mobley v. Workday, Inc., filed in the Northern District of California and allowed to proceed on an amended complaint in 2024. The plaintiff alleged that Workday’s applicant-screening algorithm functioned as an ‘agent’ of the hiring employers, making the software vendor itself liable under Title VII and the ADEA for disparate impact discrimination.

Courts had traditionally treated software vendors as tool providers, insulated from direct discrimination claims because they lacked an employment relationship with rejected applicants. The Mobley ruling cracked that assumption. If a vendor’s algorithm performs a delegated employment function, agency liability attaches regardless of who signed the offer letter.

That single procedural holding reshaped indemnification negotiations across the HR technology sector. Vendors now demand contractual liability shifting; employers now demand audit rights they never required before.

Unmonitored deployment of third-party algorithmic tools has quietly become one of the largest unbooked liabilities on corporate balance sheets, and most compliance departments have no internal mechanism to even quantify the exposure. Independent legal reference material tracking these jurisdictional shifts as they happen, such as the ongoing regulatory tracking published through the Corporate Compliance Toolkit, has become a reasonable starting point for counsel attempting to map exposure across states without commissioning a six-figure outside audit for every jurisdictional change.

Disparate Impact Doctrine Meets Machine Learning

Disparate impact liability under Griggs v. Duke Power Co., 401 U.S. 424 (1971), was built for facially neutral human policies producing discriminatory statistical outcomes. Applying that doctrine to a machine learning model trained on historical hiring data raises a harder causal question. Did the algorithm cause the disparity, or did it merely reproduce a disparity already embedded in decades of prior human decisions fed into the training set?

Plaintiffs’ counsel increasingly argue the distinction is irrelevant to liability. Courts have been receptive. The causal chain the statute cares about runs from decision to outcome, not from intent to design.

Statistical Proof Standards Now in Dispute

The EEOC’s 2023 settlement with iTutorGroup, resolving allegations that its recruiting software auto-rejected applicants over 55 and 60, established an $365,000 consent decree without a full merits trial. That settlement has become a benchmark figure cited in demand letters nationwide, despite carrying no binding precedential weight.

Case / Action	Year	Core Theory	Outcome
EEOC v. iTutorGroup	2023	Age-based algorithmic rejection	$365,000 settlement
Mobley v. Workday	2024-25	Vendor agency liability	Amended complaint sustained
NYC LL 144 enforcement actions	2023-26	Missing bias audit disclosure	Escalating per-day penalties

Insurance Markets Are Repricing the Risk Faster Than Regulators Are Writing Rules

Employment practices liability insurers have begun inserting algorithmic decision-making exclusions or sublimits into renewal policies. Underwriters cannot yet model loss frequency with confidence, so they price conservatively, which pushes premiums upward for any insured disclosing AI-assisted hiring tools on renewal applications.

That pricing shift, quietly, has become its own compliance forcing function. Boards that ignored algorithmic audit recommendations from internal counsel now face the same demand from their broker.

Documentation as the Only Reliable Defense

Across every jurisdiction reviewed, one pattern holds. Employers who can produce contemporaneous bias-testing documentation, vendor indemnification language, and a written adverse-impact review protocol fare dramatically better at the motion-to-dismiss stage than employers relying solely on vendor assurances.

Written protocol beats verbal assurance. Every time. Courts do not credit intent; they credit paper trails.

A Short Checklist Now Standard in Outside Counsel Memos

Bias audit conducted within twelve months. Vendor contract containing indemnification for discriminatory output. Adverse-impact ratio calculated and retained per role category. Human override option documented for every automated rejection.

None of that is complicated. Almost none of it was standard practice three years ago.

What the Rest of 2026 Likely Brings

Additional state legislatures, including Texas and Virginia, have bills pending that mirror the Colorado structure with narrower carve-outs for small employers. Litigation funders have begun backing coordinated multi-plaintiff filings against major applicant-tracking-system vendors, betting that Mobley-style agency theory scales.

The federal picture remains unsettled. Any executive attempt to preempt state AI statutes through agency rulemaking will run directly into the same Loper Bright doctrine that stripped agencies of interpretive deference in the first place. That structural irony is not lost on the litigators building these cases right now.