Three years in, the supply chain is shrinking, the timeline is slipping, and patients are turning to the black market. Washington’s fix isn’t working.

This is the third installment in Riptide’s ongoing coverage of the Adderall shortage. Part One examined the structural origins of the crisis. Part Two covered the DEA’s January quota increase and its limits. This update reports on what has changed — and what has gotten worse — since then.

In October 2025, the Drug Enforcement Administration announced what sounded like a turning point. The agency raised its aggregate production quota for d-amphetamine from 21.2 million grams to 26.5 million grams — a 25 percent increase, the first significant adjustment in years. The announcement was treated as a signal that Washington had finally acknowledged what patients had been living with since October 2022: a shortage that had no end in sight.

That signal was premature.

As of May 2026, the American Society of Health-System Pharmacists still lists both immediate-release and extended-release amphetamine mixed salts in active shortage. The FDA’s original shortage notice, first posted in October 2022, remains live. The shortage has now entered its fourth year. And the picture that has emerged since January is not one of a system slowly catching up — it is one of a supply chain quietly contracting around a problem that regulators have not fully solved.

The Quota Increase and the Gap Behind It

A production quota increase is not a medication increase. It is permission — permission for manufacturers to produce more, contingent on their ability to acquire the raw materials, schedule the production runs, and move finished product through a distribution system still governed by historical allocation formulas.

The gap between what the DEA authorized and what patients can actually find at a pharmacy has not closed on anything close to the timeline implied by the October announcement. Experts now estimate it could take until late 2026 or 2027 before supply consistently meets demand nationwide. That is a two-to-four-year lag between the DEA recognizing the problem and American patients reliably being able to fill their prescriptions.

Part of the delay is structural. Pharmaceutical manufacturers source the active pharmaceutical ingredients — the APIs that form the chemical basis of amphetamine medications — through global supply chains that are themselves subject to bottlenecks. A delay at a single overseas plant can cascade through multiple domestic manufacturers simultaneously. Raising the quota does nothing to accelerate that upstream supply.

Part of it is the distribution system itself. Drug distributors allocate controlled substances to pharmacies based on historical ordering patterns. A pharmacy that has seen a recent spike in ADHD patients — or one that is simply newer — may receive allocations that do not reflect its actual patient load. The quota increase flows into a distribution architecture that was not designed for the demand environment it is now navigating.

The Generic Market Is Shrinking, Not Growing

The most underreported development since January is not the pace of the quota increase. It is what has happened to the generic manufacturer base.

Mylan and Zydus have both discontinued their immediate-release amphetamine mixed salt tablets. Solco has also exited the market. Sandoz remains in shortage due to increased demand. Sun Pharma and Sandoz have discontinued their extended-release capsule lines. What remains on the ASHP’s available manufacturer list for extended-release is a short one: Shire’s branded Adderall XR, Lannett, Aytu BioPharma’s Adzenys XR ODT, and Tris Pharma’s Dynavel XR.

This is the opposite of what a functioning market response to a shortage looks like. In a normal supply-demand scenario, a persistent shortage creates an incentive for new entrants and expanded production from existing players. What has happened instead in the generic amphetamine market is consolidation through exit. Companies are leaving, not arriving.

The reasons are not hard to identify. Manufacturing Schedule II controlled substances requires DEA registration, specialized facilities, tight regulatory compliance, and exposure to quota ceilings that can limit return on investment before a production line ever reaches full capacity. The regulatory architecture that was meant to prevent diversion is also, structurally, a barrier to the market response that patients need.

The doses hit hardest by this contraction are adult doses. The 20 mg and 30 mg immediate-release tablets — the most commonly prescribed strengths for adults — remain the hardest to locate. For extended-release, the 15 mg, 25 mg, and 30 mg capsules have been the most consistently in short supply. The patients most systematically unable to fill their prescriptions are working-age adults.

The Harm the Shortage Is Producing

When the system fails to reliably provide medication, patients do not simply go without. They adapt — and some of those adaptations carry their own serious risks.

The CDC issued health advisories in 2024 and 2025 flagging a pattern that has emerged directly from the shortage: increased counterfeit pill circulation among patients who cannot obtain legitimate prescriptions. Counterfeit amphetamine products are frequently pressed with fentanyl or other adulterants. The patient who has exhausted 25 pharmacies, as one documented case involved a 22-year-old college student, faces a choice between going unmedicated and seeking supply through channels the government cannot regulate.

This is the harm calculus that the DEA’s quota-setting framework does not appear to weigh. The regulatory system was designed around the risk of diversion — the concern that legitimate prescription channels would be exploited to produce a recreational supply. What it did not adequately model is the risk created by the shortage itself: that patients cut off from the legitimate supply chain would turn to an unregulated one.

The CDC’s advisory noted particular concern about injury risk and overdose vulnerability among patients experiencing disrupted stimulant access. For patients managing ADHD, abrupt discontinuation also produces a recognized withdrawal syndrome — fatigue, depressed mood, difficulty concentrating, sleep disruption — that the clinical literature classifies as stimulant discontinuation, distinct from addiction, but real and disruptive nonetheless.

The Telehealth Variable

One factor sustaining elevated demand deserves specific attention. The DEA’s pandemic-era waiver, which suspended the Ryan Haight Act’s requirement for an in-person evaluation before prescribing controlled substances, has been extended through the end of 2026.

That waiver was a reasonable public health accommodation during COVID-19. Its continued extension reflects both the political difficulty of rescinding it and the genuine access benefit it provides to patients in areas with limited psychiatric or primary care capacity. But it also means that the demand-side pressure that contributed to the original shortage has not been reduced. Every month the waiver remains in effect, the prescription volume the supply chain must serve continues at post-pandemic levels.

This is not an argument for ending the waiver — the patients it serves are real, and abrupt rescission would create its own access crisis. It is an observation that the demand-side and supply-side policy levers are not being managed in coordination. The DEA extended prescription access while simultaneously constraining production. The results have been predictable.

What Would Actually Fix It

The structural diagnosis has not changed since this series began. The DEA’s quota system is backward-looking by design — it calculates allowable production from historical demand data, in a market where demand has undergone a sustained structural shift. The system cannot respond in real time because it was not built to.

Legislative discussions on reforming quota-setting have not produced major legislation as of May 2026. What has changed is that the DEA’s willingness to make significant quota adjustments — the October 2025 increase was the largest in years — suggests the agency recognizes the system is undersupplying a legitimate medical need.

But recognition and resolution are not the same thing. The manufacturer exits since January represent a structural deterioration that a quota increase alone cannot reverse. Rebuilding generic manufacturing capacity for Schedule II substances requires investment decisions by private companies operating under regulatory constraints that have not materially changed. The companies that left the market did not leave because of the quota ceiling alone — they left because the economics of producing tightly regulated, low-margin generics under production caps that can shift administratively did not pencil out.

Until that equation changes — through quota reform, liability adjustments, API supply chain investment, or some combination — the shortage is not a problem being solved. It is a condition being managed. And right now, it is not being managed well.

Sources: Amphetamine mixed salts shortage data and manufacturer status from the American Society of Health-System Pharmacists Drug Shortage database (ashp.org/drug-shortages), updated continuously. DEA aggregate production quota figures from the agency’s annual quota orders published in the Federal Register, including the October 2025 order setting d-amphetamine at 26.5 million grams. FDA drug shortage notice for amphetamine mixed salts, originally posted October 2022, status current as of May 2026 (accessdata.fda.gov). Supply recovery timeline projections drawn from reporting by pharmacy trade and health policy press citing pharmaceutical supply chain analysts. CDC health advisories on counterfeit stimulant pill circulation issued 2024–2025 (cdc.gov). DEA telehealth waiver extension through December 2026 published via agency rulemaking notice. Ryan Haight Online Pharmacy Consumer Protection Act, 21 U.S.C. § 831.

The federal government has a position on your rotisserie chicken.

If you use SNAP benefits — the program formerly known as food stamps — you cannot buy one. The chicken sitting under the heat lamp at your grocery store deli, the one that costs less per pound than the raw bird on the shelf six feet away, is ineligible. The USDA prohibits the purchase of “foods that are hot at the point of sale.” It doesn’t matter that the chicken is nutritionally identical to one you’d cook at home. It doesn’t matter that cooking it yourself requires a functioning stove, cookware, storage space, time, and energy, that many SNAP recipients don’t have in reliable supply. The rule is the rule.

The same program that bars a hot rotisserie chicken has, until recently, fully subsidized Mountain Dew and Skittles.

That’s not a talking point. That’s the policy. And the current push to fix it by restricting junk food — without touching the hot food ban — reveals exactly how little this debate has ever been about nutrition.

The reform wave looks decisive on paper. The Trump administration’s Make America Healthy Again initiative, led jointly by Agriculture Secretary Brooke Rollins and HHS Secretary Robert F. Kennedy Jr., has approved SNAP food-restriction waivers for 22 states as of late April 2026. Texas became the first to implement full restrictions in April. Alabama’s governor signed similar legislation two weeks ago. A national restriction on soft drinks, candy, and prepared desserts — Senate Bill 561, the Healthy SNAP Act — is pending in Congress alongside proposed Farm Bill amendments that could extend the restrictions nationwide.

The political coalition behind this is real and bipartisan at the edges. The argument is straightforward: taxpayers shouldn’t subsidize products that cause obesity, diabetes, and chronic disease, particularly in low-income communities that already bear a disproportionate health burden.

The problem is that the argument is straightforward because it leaves out most of the situation.

SNAP benefits average roughly $6 per person per day. That’s not a comfortable food budget. It’s a constraint that shapes every purchase decision in ways the junk food debate consistently ignores.

Fresh produce spoils. Lean protein is expensive. Cooking from scratch requires time, equipment, refrigeration, and physical energy — none of which are guaranteed for someone managing poverty, a disability, a second job, or children. Processed food is calorie-dense, shelf-stable, and forgiving of difficult circumstances. Choosing a bag of chips over a head of broccoli is often not a failure of nutritional knowledge. It’s rational behavior under real constraints.

The cheapest bread at the low end of the market isn’t reliably shelf-stable anymore, either. Product quality at the bottom price tier has deteriorated — lower-grade preservatives, ingredient substitutions, thinner margins — meaning the “just buy staples” answer fails even on its own terms. A two-dollar loaf can mold in days. A bag of chips lasts.

The hot food ban remains untouched in this reform wave. You can’t buy a rotisserie chicken. No state has filed a waiver to fix that. None is being encouraged to.

Meanwhile, the hot food ban remains untouched in this reform wave. You still cannot use SNAP to buy that rotisserie chicken. You cannot buy a hot deli sandwich. You cannot buy prepared soup. You cannot buy a precooked meal. The Southern Poverty Law Center — which still does some good — in a March letter to Congress, made the obvious point: instead of restricting what SNAP recipients can purchase, policymakers should be expanding eligibility to include hot foods, meal kits, and prepared grocery items. A disability advocacy organization, CLASP, has documented the specific burden the hot food ban places on people with physical disabilities for whom cooking a raw chicken represents a genuine physical challenge.

No state has filed a waiver to fix that. None is being encouraged to.

The evidence base for the restrictions is thin. ABC News medical correspondent Dr. Darien Sutton put it plainly when the first wave of waivers was announced: there is no evidence that restricting soda access will reduce chronic disease. A physician quoted by Newsweek added the constraint the whole debate sidesteps — approval may be granted, but real food costs more money.

The Congressional Budget Office has projected roughly $41 billion in federal SNAP cuts under current reconciliation legislation — the so-called Big Beautiful Bill — with about $7 billion in direct benefit reductions and another $35 billion shifted to states. Those cuts are happening simultaneously with restrictions on what the reduced benefits can buy.

The National Grocers Association and the National Association of Convenience Stores, in a joint letter to Congress, asked lawmakers to pause and evaluate the state waiver pilots before moving to national restrictions. Their concern is partly self-interested — SNAP sales account for a significant share of total US grocery revenue — but their substantive point stands: the definitions being used in proposed legislation are vague enough to create serious compliance problems at the point of sale, and no one has yet measured whether the state restrictions are producing the health outcomes their proponents promised.

The rights argument — that restricting what poor people can eat with government assistance is a form of paternalism not applied to other government programs — is legitimate and largely absent from the coverage. People receiving the mortgage interest deduction are not told what kind of home to buy. People receiving Medicare are not told what kind of food they must eat to qualify. The application of behavioral conditions specifically to poverty assistance, and specifically to food, reflects a set of assumptions about poor people that don’t survive contact with the demographics of who actually uses SNAP.

More than 40 million Americans receive SNAP benefits. Many of them are working. Many of them are disabled. Many of them are children. The policy architecture being built assumes that what they’re buying is the problem and that the solution is restriction. It does not assume that $6 a day is inadequate. It does not assume that a hot rotisserie chicken is more nutritious than the items being banned. It does not address food deserts (not desserts), where the nearest grocery store may be miles away, and the accessible option is a gas station.

The rotisserie chicken rule has been federal policy for decades, through administrations of both parties. The current administration has chosen to fight the nutrition battle by adding restrictions on soda rather than lifting the restrictions on cooked protein. That choice reveals the priorities at work.

It’s easier to ban something than to fund something. And it’s easier to restrict what poor people buy than to ask why $6 a day was ever considered sufficient to eat well.

Sources: USDA Food and Nutrition Service, SNAP Food Restriction Waiver tracker (updated April 22, 2026); U.S. Senate Bill 561, Healthy SNAP Act of 2025, 119th Congress; H.R. 2512, Hot Foods Act of 2025, 119th Congress; USDA press release, “Secretary Rollins Signs Six New State Waivers,” December 10, 2025; Congressional Budget Office scoring of HR 1 SNAP provisions; National Grocers Association letter to Congress, March 2026; Southern Poverty Law Center letter to Congress, March 2, 2026; Center for Law and Social Policy (CLASP), “The SNAP Hot Foods Ban Is Inequitable and Should Be Removed,” May 2023; ABC News, “6 More States Approved for Waivers to Remove Unhealthy Foods from SNAP Benefits,” December 12, 2025; Newsweek, “SNAP Benefit Map Shows States With Junk Food Bans in 2026,” January 2, 2026; Food Navigator USA, “Farm Bill Leaves SNAP Rules Unchanged for Now as Amendment Fight Looms,” April 30, 2026; Propel, “Food Stamps and Junk Food Bans: What You Need to Know,” March 13, 2026.

Every May 1st, labor gets its moment. Unions march. Politicians invoke the dignity of work. The governor issues a statement.

None of it will mention Idris Idowu.

Idowu is a direct support professional — a caregiver for Marylanders with developmental disabilities, the kind of worker who helps someone bathe, eat, communicate, and navigate a world that was not designed for them. He testified in Annapolis earlier this year as the General Assembly debated a second consecutive round of cuts to the Developmental Disabilities Administration, the state agency that funds his work and the work of thousands like him.

“I have a family of four depending on me,” he told rallygoers. “My children look to me to keep a roof over their heads or food on the table. Now you are proposing wage cuts — let me be clear about what that means for people like me. It means choosing between groceries and electricity.”

The legislature went ahead anyway.

They ran the same playbook. They moved the number. They blocked the most damaging individual caps. They called it a partial victory.

In April, Gov. Wes Moore signed Senate Bill 282, Maryland’s fiscal year 2027 budget, into law. Buried inside its $70.8 billion in spending was $126 million in cuts to the DDA — the second consecutive year the agency has been reduced. The year before, $164 million was cut after advocates spent months in Annapolis fighting a proposal that would have been three times worse. This year, they ran the same playbook. They moved the number. They blocked the most damaging individual caps. They called it a partial victory.

The cumulative damage from enacted state cuts is approaching $290 million over two budget cycles. Because DDA services are funded through Medicaid waivers — matched dollar-for-dollar by the federal government — the real-world impact is closer to twice that figure in total lost resources.

What that looks like on the ground is not an abstraction. It looks like reduced wages for the workers who provide one-on-one support to people with the most intensive needs. It looks like reimbursement rate cuts to community providers already operating under financial strain. It looks like the exit of experienced caregivers from a field that cannot afford to lose them.

Maryland already faces a projected shortage of 38,000 behavioral health workers by 2028. The budget that just passed makes that problem worse, not better.

Direct support professionals occupy a peculiar position in the American labor landscape. The work they do is irreplaceable — there is no automation coming for the job of supporting a nonverbal adult with complex medical needs through a morning routine — and yet the field has historically paid poverty-adjacent wages, offered minimal benefits, and suffered turnover rates that would be considered a crisis in any other sector.

In recent years, Maryland had begun to correct this. The state adjusted DDA reimbursement rates specifically so providers could raise wages and reduce the churn that destabilizes care. The investment was working. Two years of cuts are eroding it.

The mechanism matters here. When the state cuts what it pays providers, and cuts what families in the self-directed care model can pay the workers they hire directly, those workers leave. Not out of disloyalty. Out of arithmetic. A guest commentary published in the Baltimore Sun during the session put it plainly: lower compensation drives workers away, reduces availability, and increases turnover — and when the workforce becomes unstable, the burden does not fall on the system. It falls on the individual.

For the families of the roughly 19,000 Marylanders who depend on DDA services, that burden lands in the most personal possible way. Some have already absorbed it. A 60-provider survey conducted during the session found more than $35 million in uncompensated care — services delivered and not yet paid for — that providers are carrying to keep clients stable. That is not a sustainable model. It is a system held together by people doing more than they are paid to do, because the alternative is that someone gets hurt.

There is a version of this story in which the Moore administration’s position is defensible. DDA spending grew by more than $920 million during this administration. Enrollment in self-directed services increased by more than 30 percent in both 2023 and 2024. The agency has consistently overrun its annual budget. State analysts described the DDA’s finances as, in their own characterization, a black box — resistant to the kind of scrutiny that responsible fiscal management requires.

These are legitimate concerns. A program that spends 42 percent above its original appropriation in a single fiscal year creates genuine pressure on a state already staring down a $1.5 billion structural deficit. The administration is not wrong that something needed to change.

But there is a difference between managing unsustainable growth and cutting the wages of caregivers who make $15 an hour. The state has moved aggressively on the latter without demonstrating, in any public-facing way, a serious plan for the former. The DDA’s financial opacity was not created by Idris Idowu. He is paying for it anyway.

May Day belongs to workers. Not just the ones with union cards and pension plans, but the ones doing the work that holds the most vulnerable members of society together — quietly, daily, for wages that do not reflect what that labor is actually worth.

Maryland’s direct support professionals showed up in Annapolis this winter. Hundreds of them, alongside the people they care for and the families who depend on them. They testified. They rallied. They moved the number.

Then the legislature voted 102-13 to pass the budget anyway, and the governor signed it.

This is who Workers’ Day is for. They are still waiting for someone to notice.

Sources: Maryland Matters, Baltimore Sun, MDBayNews, FOX45/Spotlight on Maryland, WYPR.

Every year, American police kill roughly 1,000 to 1,300 people. Journalists, researchers, and advocates have spent a decade building databases to track who dies and why — because the federal government won’t do it systematically. Race. Age. Whether the person was armed. Whether they were fleeing.

But one variable is almost never recorded: whether the person who died had a disability.

That gap isn’t accidental. No federal law requires it. No state does it consistently. And the result is that one of the most documented patterns in American policing — disabled people dying at wildly disproportionate rates — exists almost entirely outside official government data.

What the unofficial numbers show

The most cited benchmark comes from the Ruderman Family Foundation, which analyzed high-profile police killings and estimated that people with disabilities account for between one-third and one-half of everyone killed by police — while representing roughly 12 to 20 percent of the general population.

The Washington Post’s Fatal Force database, which has tracked every fatal police shooting since 2015, finds roughly 25 percent of victims showed signs of mental illness or were in behavioral crisis. Other peer-reviewed analyses put the share higher, particularly when autism and intellectual disabilities are included alongside psychiatric conditions.

The Treatment Advocacy Center, which focuses on serious mental illness, estimates that people with untreated psychiatric conditions face roughly 16 times the risk of being killed in a law enforcement encounter compared to the general public.

For autistic people specifically, research suggests police contact rates run approximately seven times higher than for neurotypical peers — driven not by violence but by behaviors that read as suspicious or non-compliant: avoiding eye contact, repetitive movement, literal interpretation of commands, refusal to drop objects during sensory overload.

These figures are not government statistics. They come from journalists counting deaths in news reports, researchers coding bodycam releases, and advocacy organizations tracking cases families bring to them. They are the best available evidence — and they exist because no official system produces anything better.

The federal gap

The FBI operates a national Use of Force data collection program, launched in 2019 after years of pressure following high-profile killings. Participation by law enforcement agencies is voluntary. Disability status of people subjected to force is not a required data field.

The result: the most comprehensive federal database on police violence in American history cannot tell you how many disabled people police killed last year. Or the year before. Or any year.

This isn’t a technical limitation. It’s a policy choice — one that has never faced serious legislative challenge at the federal level.

What the cases look like

The pattern shows up in cases across the country, repeatedly, with similar mechanics.

In 2020, a Utah mother called police for help with her 13-year-old autistic son, Linden Cameron, who was having a sensory crisis. Officers shot him — unarmed — during a welfare check. He survived.

In 2024, Ryan Gainer, an autistic teenager in California, was shot after allegedly chasing an officer with a garden tool during what his family described as a meltdown.

On March 1, 2026, Alexander LaMorie — a 25-year-old autistic man living in a Columbia, Maryland apartment complex designed for people with disabilities — called 911 himself to report an extortion scam. Officers responding to the call shot him at least nine times after he refused to drop a knife. The entire encounter lasted under two minutes.

In each case, the person who died or was shot had called for help, or had help called for them. In each case, the encounter escalated within minutes over an object and non-compliance. In none of these cases was disability status entered into any mandatory government database as a factor in the use of force.

Why the pattern persists

Researchers and advocates broadly agree this isn’t about police targeting disabled people. It’s about structural mismatch.

Mental health crises, autism meltdowns, and episodes tied to intellectual disabilities generate 911 calls — for wellness checks, disturbances, behavioral emergencies. Police arrive trained for threat assessment and command compliance. The populations most likely to be in crisis are also the populations least able to comply with rapid verbal commands under stress.

Crisis Intervention Training exists and shows some effectiveness. Co-response programs — pairing mental health clinicians with officers — have produced promising results in pilot cities. Neither is standard. Neither is required.

And without mandatory data collection, there is no official mechanism to measure whether anything is getting better or worse.

The accountability gap

Advocates have pushed for years for federal legislation requiring disability status tracking in use-of-force reporting. It hasn’t passed. At the state level, requirements are inconsistent at best — most states, including Maryland, have no mandatory disability field in their use-of-force reporting forms.

The Autism Society of Maryland called LaMorie’s death “profoundly tragic.” Howard County officials expressed grief and cooperated with the state’s Independent Investigations Division, which is reviewing the shooting. The investigation is ongoing.

What won’t happen automatically, when that investigation concludes: any official count of how many times in Maryland — or America — a disabled person called police for help and ended up dead.

That number exists. Someone just has to count it themselves.

Alexander LaMorie’s death is under investigation by the Maryland Attorney General’s Independent Investigations Division. Bodycam footage was released on March 30, 2026.

SB 282 is now law. For the second straight year, Gov. Moore’s administration has cut the Developmental Disabilities Administration — and this time, a last-ditch effort to restore funding failed by a lopsided margin.

ANNAPOLIS — It is official. Senate Bill 282, Maryland’s fiscal year 2027 budget, passed both chambers on Monday and has been sent to Gov. Wes Moore for his signature — carrying with it $126 million in cuts to the Developmental Disabilities Administration, the state agency that funds care for more than 19,000 Marylanders living with intellectual and developmental disabilities.

The final vote was not close. The Senate adopted the conference committee report 38-6. The House followed, 102-13. A floor amendment by Del. Lauren Arikan (R-Harford) that would have restored a portion of the DDA funding failed 23-97 — the scoreboard that appears in the viral photo circulating among disability advocates this week, showing the House vote board lit up in a sea of green “yes” votes beside the words “Third Reading Passed.”

For families and caregivers who spent weeks rallying in Annapolis, testifying before committees, and flooding lawmakers with calls and emails, the overwhelming margins carry a particular sting.

“We don’t want to have to keep coming here every year,” said Carmen Houston-Ludlam of Calvert County at a rally on the State House grounds last Tuesday. The cuts, she said, are “destroying” services for the people who need them most.

This is the second consecutive year Moore’s administration has cut the DDA. Last year’s session ended with $164 million in enacted reductions, down from an original proposal of $457 million after sustained public pressure forced significant restoration. This year, the starting number was $150 million; the Senate reduced it to $126 million, and that figure held through conference.

What Is the DDA — and What Does It Do?

The Maryland Department of Health’s Developmental Disabilities Administration is the primary state agency responsible for funding community-based services for residents with intellectual and developmental disabilities — conditions including autism, Down syndrome, cerebral palsy, and related disorders. Its core mission is to connect Marylanders with the supports they need to live, work, and participate in their communities rather than in institutions.

The DDA does this primarily through Medicaid waivers: federally matched funding programs that allow the state to cover a broad range of services beyond standard Medicaid. In fiscal year 2025, approximately 19,100 people received services through those waivers. About 15,200 were enrolled with community providers — established organizations that deliver disability care — while roughly 3,900 chose the self-directed model, in which individuals or their families hire caregivers directly.

Those services span a wide range of essential supports:

• Residential supports, including live-in or group home caregiving

• Day habilitation and employment services — structured daytime programs, supported employment, and career exploration

• Transportation assistance

• Respite care — temporary relief for family caregivers

• One-on-one direct support staffing for those with the highest medical and behavioral needs

• Medical day care, assistive technology, and community development services

The agency operates on approximately $1.3 billion in combined state and federal funds for fiscal 2027. That federal partnership is central to understanding why DDA cuts are more damaging than the headline number suggests: the waivers are funded on a dollar-for-dollar match with federal Medicaid dollars. A $126 million state cut triggers a corresponding loss of roughly $126 million in federal funds, meaning the real-world impact on services is closer to $252 million.

What Is Actually Being Cut

Moore’s original fiscal 2027 proposal framed the reductions as “cost containment” — $150 million in savings achieved through policy changes rather than direct program eliminations. The Senate Budget and Taxation Committee reduced the figure to $126 million and rejected the most severe provision: a proposed $500,000 annual cap on individual care budgets that advocates warned would have forced the highest-need individuals into institutional placements.

What survived into the enacted budget:

• Caregiver wage reductions. Under the self-directed care model, individuals and families can currently set caregiver wages to reflect the intensity of care needed — known as “reasonable and customary” wages. The enacted budget lowers those wages toward Bureau of Labor Statistics benchmarks for similar jobs, with nonfamily employees receiving a slightly higher allowance than family members. The wage boost for unlicensed providers — such as camp counselors and art instructors who currently provide services under self-directed plans — is eliminated entirely.

• A 2% cut to provider reimbursement rates across all community services. Laura Howell, CEO of the Maryland Association of Community Services, noted that most of the services subject to that cut are currently funded below 100 percent of actual cost. “That’s going to be painful. There’s no question about it,” she said.

• Reduction in Low-Intensity Support Services (LISS) from $5.5 million to $2 million. The program helps families afford one-time purchases or small accommodations — adaptive equipment, home modifications — for family members with disabilities. Moore’s original proposal eliminated it entirely; last year’s session cut it and then restored it. It survives again, but at a significantly reduced level.

Who Gets Hurt

The cuts fall hardest on those with the most intensive needs — individuals who require dedicated one-on-one staffing, families who rely on self-directed care because no community provider can accommodate their loved one’s medical complexity, and the caregivers themselves, whose wages are now being directly squeezed by state policy.

Tracie Feron, a Baltimore County resident, is the mother of 30-year-old Connor, who has autism and complex medical conditions requiring around-the-clock care. She told Maryland Matters that residential facilities have already rejected Connor because they cannot meet his needs. His DDA waiver lets the family hire support staff directly. The enacted wage reductions put those workers — and Connor’s stability — at risk.

Advocates warn that even modest wage reductions in a field already plagued by high turnover and low pay will accelerate a worker exodus. If caregivers leave, there may be no one to replace them. And the alternative — institutional placement — would cost the state far more in the long run, while removing individuals from the communities and relationships they have built.

Baltimore Today reported that the changes could reduce home-based caregiver wages by as much as 50 percent for some workers in self-directed arrangements. Former Baltimore Orioles player BJ Surhoff, whose son has autism, called the cuts “reprehensible” at last week’s rally. His presence drew attention, but advocates were clear: the names filling the plaza are not the point. The names on the caregiving rosters are.

Angela Noppenberger, a mother of two teenage children with developmental disabilities and a longtime Baltimore City social worker, described the DDA’s existing appeals process as “very stressful and taxing on families” — even for someone with her professional background. The new cuts, she told legislators, will make that already difficult system harder to navigate for the families least equipped to fight it.

The Vote: What the Margins Mean

Senate Bill 282 moved through the General Assembly on a fast-tracked but contested path. After the Senate passed the budget 40-6 on March 18, the House Appropriations Committee made amendments the Senate refused to accept, triggering a conference committee. The conferees — led by Senate Budget Chair Guy Guzzone (D-Howard) and House Appropriations Chair Ben Barnes (D-Anne Arundel and Prince George’s) on their respective sides — reached agreement over the weekend.

Both chambers adopted the conference report on Monday, March 30. The Senate voted 38-6. The House voted 102-13.

On third reading in the House, Del. Arikan’s floor amendment — which would have restored a portion of the DDA cuts — failed 23-97. The amendment was among several Republican-sponsored floor proposals that were voted down over the course of the budget debate; an earlier amendment by Del. Hartman (R) was also rejected, 32-97, and a proposal by Del. Chisholm (D) failed 36-97.

The overwhelming final passage votes — bipartisan margins in both chambers — make clear that no realistic path to restoring the DDA cuts existed once the conference report was agreed to. Disability advocates who packed the State House over the past several weeks succeeded in moving the number from $150 million to $126 million and blocking the $500,000 individual cap, but could not stop the enacted reductions.

Senate President Bill Ferguson acknowledged the scale of DDA spending growth as the driver of the decision. “We have to look at the place where the costs are escalating the fastest, and that is within the DDA budget,” he said. “Moving forward, we cannot afford to have those levels of increases.”

The Budget Backdrop: A ‘Black Box’ Agency and Runaway Costs

State officials argue the cuts are necessary because DDA spending has grown to genuinely unsustainable levels. Since 2021, the agency has consistently overspent its annual budget — by $450 million in fiscal 2025 alone, roughly 42 percent above original appropriations. Enrollment growth in self-directed services was particularly sharp: more than 30 percent in both 2023 and 2024, driven largely by young adults transitioning out of the school system and into Medicaid waiver programs.

State budget analysts have been candid about not fully understanding what is driving the cost increases. They have described the DDA budget historically as “a black box” resistant to outside scrutiny. A recent legislative audit compounded the concern, finding the agency had failed to recover nearly $119 million in advance payments made to providers during a troubled transition to a new fee-for-service payment system, with more than 100 providers still owing money.

Health Secretary Meena Seshamani told legislators the proposed cost-containment measures are necessary to keep the waiver program financially viable and to maintain federal approval for it to continue. The Centers for Medicare and Medicaid Services must approve changes to Medicaid waiver programs, adding a layer of federal review that could further complicate implementation.

But advocates offer a counternarrative: rising costs reflect a success. People with developmental disabilities are living longer and healthier lives precisely because they are living in the community rather than institutions. “That is fantastic, but it does mean that it costs more,” said Howell. The DDA’s own enrollment data supports this: Maryland is serving more people, with more complex needs, than at any point in the agency’s history.

Guzzone said the state will be adding chief financial officer capacity within the DDA and hiring additional fiscal consultants in an attempt to stabilize the agency’s finances — improvements advocates have long sought. Whether those changes come fast enough to prevent a third consecutive year of cuts remains to be seen.

Two Years Down. The Disability Community Looks Ahead.

Last year’s session began with a $457 million DDA cut proposal so severe that advocates called it potentially catastrophic. Hundreds of people rallied in January cold. Weeks of testimony, organizing, and political pressure ultimately restored $300 million of those proposed cuts — but $164 million was enacted.

This year, the community returned to Annapolis and ran the same playbook. They moved the number from $150 million to $126 million and blocked the individual budget cap. Those are real wins. But $126 million — compounded onto $164 million from last year — is a cumulative wound.

Rachel London, executive director of the Maryland Developmental Disabilities Council, was clear-eyed about what remains after the session’s budget framework was announced: “The $457 million in general and federal funds that were cuts in fiscal year 2026 still remain, and that means we have to continue our advocacy for that funding.”

Alicia Wopat, president of the Self-Directed Advocacy Network of Maryland, had warned at the start of this session that “these proposals will have severe consequences.” With the bill now passed and headed to the governor’s desk, the community is beginning to assess exactly what those consequences will look like on the ground — in group homes, in day programs, in the homes of families who have built their lives around the stability of DDA support.

For roughly 19,000 Marylanders who depend on those supports, the message coming out of Annapolis two years running is the same: more will be asked of them, and less will be given.

Editor’s note: MDBayNews will continue to follow the DDA story through implementation and into the FY2028 budget cycle. Readers with direct experience with DDA services are encouraged to contact us to share their story.

A rare bipartisan backlash is building across Maryland as counties push back against proposed cuts to disability services under the administration of Wes Moore.

In letters, testimony, and direct appeals to state leadership, officials from Montgomery County, Baltimore County, Cecil County, and Wicomico County are warning that proposed cuts to the Maryland Developmental Disabilities Administration (DDA) could have devastating consequences for families who depend on these services.

At issue is a proposed $150 million reduction in DDA funding as part of Maryland’s Fiscal Year 2027 budget — a move that local leaders say could balloon into more than $300 million in real impact once federal matching funds are lost.

But behind the numbers lies a deeper question:

Why is Maryland choosing to cut care for the disabled while continuing to expand spending elsewhere?

A Budget Built on the Vulnerable

The DDA is not a discretionary program.

It funds basic life support for individuals with developmental disabilities — housing, job placement, transportation, and daily care. These are not luxuries. They are the difference between stability and crisis.

Yet as Maryland faces a projected $1.5 billion budget shortfall, the Moore administration has placed these services on the chopping block.

Local officials are now openly questioning that decision.

In a March 20 letter, leaders in Cecil County warned that cuts of this magnitude could “place additional strain on service providers and caregivers,” threatening the stability of already fragile support systems.

Wicomico County echoed the same concern, signaling that this is not an isolated complaint but part of a broader pattern.

Montgomery County Sounds the Alarm

Even in deep-blue Montgomery County — a jurisdiction that typically aligns with state leadership — officials raised serious concerns.

Their testimony outlines a cascading set of consequences:

• Over $300 million in total funding loss when federal dollars are included

• Caps on individual care budgets

• Increased administrative burdens on providers

• Reduced access to services for high-needs individuals

This is not trimming excess.

This is restructuring the system in ways that could push the most vulnerable residents out of it entirely.

Baltimore County Escalates the Fight

In Baltimore County, officials took the unusual step of directing their concerns straight to legislative leadership.

Their warning was blunt:

Cuts of this scale will lead to delays, reduced responsiveness, and a system increasingly unable to meet demand.

They also noted that this is not a one-time proposal. The administration has attempted similar cuts in prior years — suggesting a pattern rather than a one-off response to fiscal pressure.

A Rare Bipartisan Break

What makes this moment politically significant is not just the policy — it’s the alignment.

Counties that rarely agree on anything are now united in opposition.

• Progressive Montgomery County

• Suburban Baltimore County

• Rural Eastern Shore jurisdictions like Wicomico

• More conservative counties like Cecil

All are sending the same message:

These cuts go too far.

The Bigger Question: Priorities

Every budget is a statement of priorities.

And this one is raising uncomfortable questions.

Maryland continues to fund expansive initiatives across climate policy, education restructuring, and administrative growth.

Yet when the budget tightens, it is disability services — among the most essential and least flexible programs — that are first on the table.

That choice is not just fiscal.

It’s political.

Real-World Consequences

For families, this is not a policy debate.

It is a question of whether services will still exist.

Many individuals served by the DDA cannot simply “adjust” to cuts. There are no private-market alternatives waiting in the wings. When services disappear, the burden shifts immediately to families, caregivers, and already overstretched local systems.

That means:

• More people going without care

• More pressure on emergency services

• More long-term costs pushed downstream

In other words, short-term savings may create long-term liabilities.

Annapolis Under Pressure

With multiple counties now formally opposing the cuts, pressure is building on both the Moore administration and legislative leadership to reconsider.

This is no longer a quiet budget adjustment.

It is becoming a defining test of leadership — and of whether Maryland’s government is willing to protect its most vulnerable residents when it matters most.

Because once these services are cut, rebuilding them is far harder than maintaining them.

And the people affected do not have the luxury of waiting for the state to get it right later.

Two years of internal state reports reveal falling admissions, collapsing releases, and a system increasingly unable to move patients out — while youth beds sit half-empty.

The numbers inside Maryland’s monthly psychiatric inpatient reports don’t scream. They accumulate. Month after month, facility after facility, the same quiet signals repeat: more patients on the books, fewer cycling out, and an occupancy crisis at the adult level that the state has not publicly addressed.

A review of nearly two full fiscal years of data — 19 monthly reports from the Maryland Department of Health, spanning February 2024 through January 2026 — reveals a system under structural pressure that predates the current fiscal year and is worsening by measurable degrees. The data spans the final months of fiscal year 2024, all of fiscal year 2025, and the first seven months of fiscal year 2026.

“Month after month, the same quiet signals repeat: more patients on the books, fewer cycling out, and a crisis the state has not publicly addressed.”

The revolving door has jammed

The most important number in these reports isn’t the occupancy rate. It’s readmissions.

Across the full fiscal year 2025 (July 2024 through June 2025), Maryland’s state psychiatric system recorded 234 readmissions — patients returning to inpatient care after a prior hospitalization. That’s down from 300 readmissions in fiscal year 2023, a 22% drop in two years.

By the nine-month mark of fiscal year 2025 — through March 2025 — readmissions had already fallen to 178 from 230 over the same period in FY2023, a drop of 52. That is the steepest year-over-year readmission decline in this dataset at any comparable reporting point.

In any other context, falling readmissions would be celebrated as evidence that patients are doing better in the community and not relapsing. But the rest of the data tells a different story. The average number of patients on the books has not fallen — it has held steady above 1,086 across both fiscal years. Net releases, the measure of how many patients successfully transition out, fell from 1,031 in FY2023 to 985 in FY2025.

The picture that emerges is not one of patients recovering and staying well in the community. It is one of patients staying in the hospital longer, so that the revolving door has fewer rotations simply because fewer people are leaving.

Table 1: System-wide trends, FY2023–FY2025. Source: Maryland Department of Health monthly statistical reports.

Where patients are going — and not going

The shift in where admissions flow within the system is striking over three fiscal years.

Eastern Shore Hospital Center has seen the steepest decline: 119 admissions in FY2023, 99 in FY2024, and just 66 in FY2025 — a 45% drop over two years. Its net releases fell in lockstep, from 121 to 67. A facility that was moving more than 10 patients per month into the community two years ago now moves fewer than 6. The nine-month FY2025 data through March confirms the trend: Eastern Shore had just 52 total admissions through March 2025, compared to 87 over the same period in FY2023 — a decline of 35. Eastern Shore ran at or above 98% average occupancy throughout, suggesting the problem is not a low-census facility winding down. It is a facility that has stopped turning over.

Spring Grove, by contrast, saw admissions hold nearly flat at the top of the system — 332 in FY2023, 363 in FY2024, 362 in FY2025 — while its net releases rose modestly from 318 to 369. Spring Grove is the one facility that appears to be moving patients at pace with its intake. But it is doing so while running above 100% of its operated capacity, peaking at 107.3% in the most recent reporting period. In July 2025 alone, Spring Grove admitted 45 patients in a single month — the highest single-month figure in this dataset — foreshadowing the overcrowding that followed through fall and winter.

Perkins Hospital Center, which primarily serves patients from the criminal courts, saw a 21% drop in admissions from FY2023 to FY2025, while its census remained essentially unchanged. Thomas B. Finan Center in Cumberland, meanwhile, is a notable outlier on the positive side: its net releases have risen from 33 in the eight-month FY2023 period to 40 in the same period of FY2025, and its admissions have grown modestly. Finan is taking on more work and moving more patients — but it is also running at 98.5% average occupancy and has peaked at 104.5%.

Table 2: Admissions and occupancy by facility, FY2023 vs. FY2025. Figures approximate from cumulative report tables.

The RICA question

While adult facilities strain, the two Regional Institutes for Children and Adolescents — RICA Baltimore and RICA Montgomery — have run at a fraction of their capacity for two full fiscal years.

RICA Baltimore has a 45-bed operated capacity. Its average daily census has ranged from 17 to 27 patients across the 19 months of this review — never above 60% of capacity, and frequently below 53%. RICA Montgomery, with 50 beds, has averaged 19 to 31 patients. Combined, the two facilities have left roughly 40 to 55 youth psychiatric beds unfilled every month, for at least 19 consecutive months.

The February 2025 data shows RICA Baltimore at an average daily census of 21 — 47% of capacity. RICA Montgomery averaged 21 as well, or 42% of capacity. By March 2025, RICA Baltimore was at 23 patients (51%) and RICA Montgomery at 20 (40%). These figures are consistent with what the reports show month after month.

The April through June 2025 reports — the end of fiscal year 2025 — show RICA Baltimore’s monthly admissions hitting zero in June. Zero admissions to a 45-bed youth psychiatric facility in a single month, in a state with well-documented adolescent mental health needs.

“Zero admissions to a 45-bed youth psychiatric facility in a single month — while adult units were full.”

This is not explained anywhere in the reports. The data does not tell us whether admissions criteria changed, whether the referral pipeline dried up, whether staffing shortages forced bed closures within the unit, or whether youth are being diverted to private or out-of-state placements at greater public expense. But the numbers are consistent and persistent.

The ‘operated capacity’ floor

Every occupancy figure in these reports is measured against “operated capacity,” defined as beds in buildings currently in use. This is not the licensed capacity of each facility — it is the number of beds the state is actually staffing and running at any given time.

The system’s total operated capacity is listed as 1,161 beds across the adult and youth facilities. When the state reports that Spring Grove is at 107% of operated capacity, it does not mean the building is physically overwhelmed. It means the state is running fewer beds than the physical plant could support, and those beds are overfull. The difference between licensed capacity and operated capacity reflects choices — about staffing levels, about funding, about which wings to open.

Those choices are not visible in these reports. MDH did not respond to a request for comment.

A death in the facility — and deaths outside it

The reports document patient deaths across the full dataset. Over the first nine months of fiscal year 2025 through March 2025, six deaths had been recorded in the system — one more than at the same point in FY2023 and FY2024.

Notably, the March 2025 report records one in-facility death at Perkins Hospital Center. This is a departure from the pattern seen across most of the dataset, in which deaths occurred while patients were on conditional release or another form of leave status. The Perkins in-facility death in March is the only confirmed in-facility death in the 19 months of data reviewed.

Over the seven additional months from July 2025 through January 2026, four more deaths were recorded — in October, November, December, and January — all at Spring Field or Spring Grove, and all while patients were on conditional release. The reports record the deaths by facility and month but provide no information on cause or circumstance.

A system holding still while pressure builds

The quietest finding in this dataset may be the most telling. Despite falling admissions and fewer readmissions, the total number of patients on the system’s books has barely budged in two years.

In March 2023, the average number of patients on hospital books systemwide was 1,096. Through March 2025, that figure stood at 1,093. In January 2026, it was 1,094. The system’s population has remained within a three-patient range — above 1,086 — for the entirety of this review period.

The end-of-month census peaked at 1,103 on the books as of March 31, 2025 — the highest end-of-month figure in the full dataset. By January 31, 2026, it stood at 1,088.

The system is not growing. But it is not releasing patients either. Its population has stabilized at a level that keeps adult facilities chronically at or above operated capacity — while the youth units sit largely unused — not through any deliberate policy anyone has announced, but through the slow arithmetic of patients staying longer and fewer moving on.

Whether that reflects appropriate clinical decisions, a failure of community mental health infrastructure to absorb discharged patients, or some combination, the state has not said. The data only shows the result.

METHODOLOGY

Data source: Maryland Department of Health Psychiatric Inpatient Facilities Statistical Reports, February 2024 through January 2026 (19 monthly reports), obtained through public records. Fiscal year data drawn from cumulative tables published in each month’s report. FY2023 baseline figures are drawn from the comparative columns in FY2024 and FY2025 reports. “Operated capacity” is as defined by MDH in each report: beds in buildings currently in use, which may differ from licensed capacity. All figures are drawn directly from official monthly reports; no figures were projected or interpolated except where noted as approximate in tables above.

For the past several years, artificial intelligence has been sold to the public as a technological miracle—an innovation that would eliminate tedious tasks, increase productivity, and free human beings to focus on creative and strategic work. Corporate executives, consultants, and technology evangelists promised a world where AI would function like an army of digital assistants, quietly handling the drudgery of modern work.

But a growing body of research suggests that the reality may be far more complicated.

A new study by Boston Consulting Group (BCG), published in the Harvard Business Review in March 2026, has introduced a striking new phrase to describe an emerging problem: “AI brain fry.” The term describes a form of acute mental fatigue caused not by doing too little work, but by supervising too many AI systems at once.

And while the concept might sound like a humorous piece of Silicon Valley jargon, the implications are serious. As businesses rush to adopt artificial intelligence tools, the human cost of managing those systems is becoming harder to ignore.

The Rise of “AI Brain Fry”

The BCG study surveyed 1,488 full-time workers across the United States, examining how employees interact with AI tools in modern workplaces. Participants ranged from engineers and marketing professionals to finance and operations managers.

The researchers found that about 14 percent of AI users reported experiencing symptoms consistent with “AI brain fry.” In industries where workers must supervise multiple AI systems—such as engineering, IT, and marketing—the number rose to more than 25 percent.

Those experiencing the phenomenon reported a range of cognitive symptoms:

• Mental fog

• Difficulty concentrating

• Slower decision-making

• Headaches

• Increased errors

• Decision fatigue

• A constant sense of mental “noise”

One engineering manager interviewed for the study described the experience as “like having a dozen browser tabs open in my head, all fighting for attention.”

Another said their thinking felt “buzzing with static.”

In other words, AI may not be removing work from the human brain—it may simply be changing the type of work the brain must do.

The Supervisory Trap

Much of the hype surrounding AI has focused on automation. The idea is simple: machines handle repetitive tasks so humans can focus on higher-level thinking.

But in practice, many workers are not being freed from tasks—they are being transformed into AI supervisors.

Instead of writing a report, for example, a worker may now manage three or four AI tools that generate drafts, analyze data, and produce summaries. Each of those outputs must be checked, corrected, and verified.

If one system produces incorrect information, the worker must intervene.

If multiple systems produce conflicting outputs, the worker must reconcile them.

And if an AI system behaves unpredictably—as they sometimes do—the worker becomes responsible for damage control.

One anecdote cited in coverage of the study described a safety director at a major technology company who had to urgently intervene when an AI system began deleting large portions of her email inbox.

The problem, researchers say, is that supervising AI agents requires constant mental switching between tasks—an activity known to increase cognitive load.

The result is a new kind of digital exhaustion.

Too Many Tools, Too Little Clarity

Perhaps the most revealing finding from the study is how productivity changes as workers add more AI tools to their workflow.

According to BCG:

• Workers using one to three AI tools generally reported increased productivity.

• Workers using four or more tools simultaneously experienced declining productivity.

The reason appears to be cognitive overload. Managing multiple AI systems requires constant attention, monitoring, and correction.

In effect, workers are juggling several semi-autonomous assistants at once—each capable of producing useful output but also capable of creating mistakes.

Imagine supervising four interns simultaneously while they all attempt to complete different tasks at high speed.

That, researchers say, is essentially what modern AI-heavy workflows resemble.

Not Burnout—But Something New

One of the more surprising findings from the study is that workers experiencing AI brain fry reported less chronic burnout overall than workers who use little or no AI.

This suggests the phenomenon may not represent long-term exhaustion but rather an acute form of cognitive strain.

In many cases, symptoms reportedly fade after workers take breaks or step away from AI-intensive tasks.

Still, the short-term effects can be significant.

Workers experiencing brain fry reported:

• 33 percent more decision fatigue

• Increased small mistakes

• Greater frustration with tasks

• Higher intentions to quit their jobs

In a labor market already strained by stress and burnout following the pandemic-era shift to remote work, the last thing employers should want is another driver of turnover.

Yet that may be exactly what poorly managed AI adoption creates.

The Workslop Problem

Another issue complicating AI adoption is the growing problem of low-quality AI output, sometimes referred to by researchers as “workslop.”

This occurs when AI tools generate large amounts of content that appear useful but ultimately require extensive human cleanup.

Instead of saving time, the tools can create additional work.

Employees must fact-check AI-generated information, rewrite poorly structured drafts, and verify data accuracy.

In some organizations, workers quietly report that the time spent correcting AI output can rival—or even exceed—the time it would have taken to do the task manually.

The result is a paradox.

AI promises efficiency, yet poorly implemented AI systems can produce the opposite effect: more tasks, more monitoring, and more cognitive strain.

A Familiar Pattern in Technology

To be clear, the emergence of “AI brain fry” does not mean artificial intelligence is inherently harmful.

Rather, it reflects a pattern seen repeatedly throughout technological history.

When email first entered workplaces in the 1990s, it was supposed to simplify communication. Instead, many workers found themselves overwhelmed by an endless stream of messages.

The same occurred with smartphones, collaboration platforms, and remote work tools. Each innovation promised efficiency but often introduced new forms of mental load.

AI may simply represent the next phase in that cycle.

As BCG researchers put it, the situation is similar to handing a Ferrari to someone who has just learned to drive.

The power is there.

But without the right systems and training, the speed can become overwhelming.

What Leaders Should Do

The Harvard Business Review study emphasizes that the problem is not AI itself—it is how organizations deploy it.

Companies eager to appear technologically advanced often add AI tools to existing workflows without redesigning the work itself.

This approach, researchers say, almost guarantees cognitive overload.

To address the problem, the study recommends several practical steps.

First, companies should redesign workflows rather than simply layering AI tools onto existing jobs. AI should be integrated thoughtfully into processes rather than bolted on as an afterthought.

Second, organizations should limit the number of AI systems employees must supervise simultaneously. Evidence suggests productivity begins declining when workers manage more than three or four tools at once.

Third, leaders should shift performance metrics away from raw output or activity levels. If employees are rewarded simply for generating more AI-driven content, they may feel pressure to juggle too many systems at once.

Fourth, companies must invest in training workers to manage AI effectively. Supervising AI agents requires new skills, including knowing when to trust outputs and when to intervene.

Finally, managers should treat human attention as a limited resource. Scheduled breaks from AI-heavy tasks, focused work blocks, and realistic expectations can reduce mental fatigue.

These are not radical ideas. They reflect the simple reality that technology should adapt to human needs—not the other way around.

The Political and Cultural Question

Beyond workplace management, the rise of AI brain fry raises broader cultural questions.

In recent years, policymakers and technology leaders have focused heavily on the economic benefits of artificial intelligence: productivity gains, automation, and new industries.

But far less attention has been paid to the psychological consequences of living and working alongside increasingly autonomous machines.

As AI systems become more integrated into everyday life—from workplace software to personal assistants—the cognitive burden of managing those systems could become a major mental health issue.

In a society already struggling with attention fragmentation, digital overload, and rising anxiety levels, adding another layer of technological complexity may carry risks.

The promise of AI has always been that machines would do more of the thinking so humans could do less.

But if workers must constantly monitor, correct, and second-guess those machines, the result may not be liberation.

It may simply be a new form of mental noise.

The Bottom Line

Artificial intelligence remains one of the most powerful technological innovations of the modern era. Properly implemented, it has the potential to transform industries, accelerate scientific discovery, and unlock new forms of productivity.

But the early evidence surrounding “AI brain fry” should serve as a warning.

Technology does not automatically make life easier.

Without thoughtful implementation, it can simply change the shape of our workload—and sometimes intensify it.

As businesses race to adopt AI tools and policymakers push for rapid innovation, the real question may not be how powerful artificial intelligence becomes.

The real question may be how much cognitive strain the human brain can handle.

If companies ignore that reality, the next workplace crisis may not come from automation replacing workers.

It may come from workers struggling to keep up with the machines meant to help them.

As the United States moves deeper into 2026, the long-running shortage of Adderall and its generic equivalents remains unresolved. While federal regulators have taken steps to modestly increase production limits for key stimulant medications, patients and pharmacies across the country continue to report inconsistent access, delayed refills, and forced medication changes.

The shortage, which began in late 2022, has become a case study in how federal regulation, supply-chain fragility, and surging demand can collide — with real consequences for millions of Americans managing ADHD.

Shortages Still Listed, Access Still Uneven

As of mid-January 2026, the Food and Drug Administration continues to list multiple formulations of amphetamine mixed salts — including immediate-release and extended-release versions of Adderall — on its Drug Shortages Database. Updates as recent as January 15 indicate limited availability or back orders across several manufacturers and dosage strengths.

Major producers such as Teva report constrained supplies of immediate-release tablets, with some estimated recoveries pushed into late January or February. Other manufacturers, including Rhodes, Amneal, and Camber, report extended back orders for certain extended-release capsules, stretching into early 2026. While some brand-name formulations and select generics remain available in limited quantities, pharmacy-level access varies dramatically by region and distributor.

The American Society of Health-System Pharmacists has echoed these findings, citing intermittent manufacturing delays, allocation limits, active ingredient issues, and sustained demand growth.

DEA Raises Production Ceiling — But It’s Not a Silver Bullet

In early January, the Drug Enforcement Administration finalized its 2026 Aggregate Production Quotas (APQs) for Schedule II controlled substances. Following more than 5,000 public comments — many from ADHD patients and advocacy groups — the agency increased the national production ceiling for d,l-amphetamine, the primary component in Adderall, by approximately 14 percent above the originally proposed level.

The quota for lisdexamfetamine, the active ingredient in Vyvanse, was raised by roughly 22 percent.

From a policy perspective, this reflects a rare acknowledgment that prior quota levels may have failed to keep pace with legitimate medical demand. The DEA cited prescription growth of roughly 6–7 percent annually for stimulant medications in recent years, alongside public concern about ongoing shortages.

However, quotas set a maximum — not a mandate. Manufacturers are not required to produce up to the limit, and higher ceilings do not immediately translate into pills on pharmacy shelves. Production lead times, ingredient sourcing, distribution bottlenecks, and corporate allocation decisions all remain outside the DEA’s direct control.

A Regulatory System Under Strain

The Adderall shortage has reignited debate over how the federal quota system operates. Under the Controlled Substances Act, the DEA is tasked with balancing legitimate medical need against diversion and abuse risks. That balance relies heavily on historical data, manufacturer forecasts, and federal tracking systems — all of which can lag behind rapid changes in diagnosis rates, telehealth prescribing, and patient demand.

Critics argue the system is inherently reactive rather than responsive, leaving patients caught between regulatory caution and real-world necessity. Supporters counter that quotas remain essential to preventing diversion and overproduction, particularly for Schedule II stimulants.

What’s increasingly clear is that quotas alone cannot fix a supply chain already strained by manufacturing delays, corporate consolidation, and uneven distribution practices.

Patients Feel the Consequences

For patients, the impact is tangible. Social media platforms and online forums remain filled with accounts of pharmacies unable to fill prescriptions, partial fills that disrupt treatment plans, and forced switches to alternative medications. Some report success finding supply after the holiday season, while others describe months-long disruptions with no clear resolution.

Medical professionals continue to urge patients to plan refills early, stay in close contact with prescribers, and remain open to alternatives — whether different formulations, other stimulant classes, or non-stimulant options — depending on availability and clinical suitability.

Incremental Progress, No Immediate End in Sight

The 2026 quota increases represent a modest but meaningful shift after years of patient advocacy. They may ease pressure gradually as manufacturers ramp up production over the course of the year. But they do not guarantee consistent access, nor do they resolve deeper structural issues in how controlled medications are produced and distributed.

For now, the Adderall shortage remains a reminder that even in a high-tech, data-driven healthcare system, policy decisions made far upstream can leave millions navigating uncertainty at the pharmacy counter — with focus, productivity, and quality of life hanging in the balance.

For more than three years, Americans with legitimate ADHD diagnoses have been stuck in a medication scavenger hunt. The Adderall shortage that began in late 2022 has dragged on through late 2025, leaving families, students, and working adults scrambling—often monthly—to find prescriptions that once were routine. This is not a mystery of capitalism run amok. It’s a predictable result of rigid regulation colliding with real-world demand.

At its core, the shortage reflects a basic failure to adapt. Demand surged. Supply didn’t. And federal policy lagged badly behind both.

Demand Exploded—Predictably

ADHD diagnoses and stimulant prescriptions have risen sharply over the past decade. Prescriptions for stimulant medications increased by roughly 60% from 2012 to 2023, driven by three major forces:

• Adult diagnosis: Millions of adults—especially women—are being diagnosed later in life as awareness improves.

• Telehealth expansion: Pandemic-era policies made evaluations and follow-ups far more accessible.

• Reduced stigma: ADHD is increasingly recognized as a real neurodevelopmental condition, not a character flaw.

None of this was sudden. All of it was measurable. Yet supply planning failed to keep pace—especially for generic medications, where profit margins are thinner and manufacturers are slower to scale.

The DEA Bottleneck

Adderall’s active ingredients are tightly controlled under Schedule II rules, with annual production quotas set by the Drug Enforcement Administration. Those quotas are intended to prevent abuse and diversion—but in practice, they have become a blunt instrument.

For years, quotas were based on backward-looking data rather than real-time need. Only in October 2025 did the DEA significantly raise limits—boosting quotas for key amphetamine ingredients by roughly 25%. That move was overdue, but it won’t produce instant relief. Manufacturing, compliance, and distribution don’t turn on a dime.

Even worse, quota increases don’t guarantee production. Manufacturers must still navigate labor constraints, supply chain delays, and financial incentives that often favor brand-name drugs over generics.

Manufacturing Isn’t the Whole Story

Major suppliers have acknowledged intermittent production issues. Teva, the largest supplier of both brand-name and generic Adderall, has cited demand spikes and operational delays. Other manufacturers have reported specific dosages on backorder into late 2025.

Meanwhile, pharmacies face their own restrictions on ordering controlled substances, creating wild regional variability. One store may have medication while another—across the street—has none. Patients are left calling pharmacy after pharmacy, often treated with suspicion despite valid prescriptions.

The American Society of Health-System Pharmacists continues to report inconsistent availability across immediate-release and extended-release formulations. The Food and Drug Administration has removed some products from its shortage list at times—but generics remain unreliable.

Who Gets Hurt

This shortage isn’t an abstraction. It hits:

• Children whose school performance depends on consistent treatment.

• Adults managing demanding jobs and families.

• Women, who are more likely to be diagnosed later and face skepticism when seeking treatment.

• Lower-income patients, who rely on generics and can’t easily switch medications or pharmacies.

Interruptions aren’t harmless. They lead to lost productivity, academic struggles, emotional distress, and—in some cases—dangerous medication gaps.

A Better Balance Is Possible

The choice is not between preventing abuse and ensuring access. A smarter system can do both.

That means:

• Dynamic quotas that respond to real-time prescription data.

• Clearer coordination between regulators and manufacturers.

• Pharmacy-level flexibility that reduces arbitrary local shortages.

• Respect for patients, not blanket suspicion.

In the meantime, patients are left to cope—asking pharmacists to check other suppliers, discussing alternatives like methylphenidate or non-stimulants with doctors, and monitoring federal shortage updates like amateur supply-chain analysts.

This is what happens when bureaucracy treats a growing medical need as an afterthought. The Adderall shortage is not a failure of medicine or demand. It’s a failure of governance—and until policy catches up to reality, Americans with ADHD will keep paying the price.