Imagine you’re at the supermarket checkout. Your cart is full. The line behind you is long. You tap your card. Declined.

You try again. Declined.

You haven’t overspent. You haven’t done anything suspicious. But somewhere inside your bank’s computer systems, a machine made a decision about you in less time than it takes to blink – and it made a mistake.

What just happened? And why does it keep happening to people who haven’t done anything wrong?

This isn’t a rare glitch, but something that happens to millions of people every day. And most of us have no idea why it happens or what we can do about it. The answer lies inside a fraud detection system powered by AI.

As a data science teaching professor and former financial-services data scientist, I understand how this system works and can explain why it sometimes fails the very customers it’s meant to protect. Just as important, I can help you find out what you need to know and what you can do if you or your loved ones are unfairly flagged.

A decision in milliseconds

When you tap your card, a signal travels to your bank’s fraud detection system in the time it takes to blink. The transaction processing at your checkout is fully automated, operating within AI systems that handle millions of payments simultaneously, and computes a risk score based on dozens of features extracted from that single moment. Those features might include the transaction amount relative to your recent spending average; the type of merchant; your geographic location; the time of day; the device used for online purchases; and how this purchase compares to your historical patterns.

Once those factors are plugged in, an algorithm scores your purchase in real time. A model trained on millions of past transactions then assigns each combination of features a probability on how likely it is that this transaction would be fraudulent. If that probability crosses a threshold, the transaction is blocked or flagged for review. The whole process takes less than 200 milliseconds.

‘99% accurate’ still fails millions of people

What sets this technology apart is speed. Financial institutions process millions of transactions every day, which is far greater than any human team can effectively monitor. Banks also have fraud analysts, but their work happens at a different layer entirely – reviewing patterns, investigating cases, and handling disputes that the automated system escalates to them.

To their credit, these new systems are usually accurate at catching fraud. Banks lose far less money due to card fraud today than they did before machine learning – one of the foundational technologies that power today’s AI systems – became standard.

Still, the word “accurate” conceals a problem. Consider the numbers. The Federal Trade Commission reported that Americans lost more than $12.5 billion to fraud in 2024 – a 25% increase from the year before. As banks process more transactions than ever, fraudsters are keeping pace, too.

And here is the part that is especially worth noting: According to Stripe, one of the world’s largest payment processors, “false declines” (legitimate transactions wrongly rejected) are a structural problem across the entire industry, and industry research consistently suggests they cost the financial system more than actual fraud does.

These errors aren’t random. They cluster around people and situations that the algorithm wasn’t properly trained to expect. Buying gas in a city you’ve never visited or making a large rent payment for the first time aren’t inherently suspicious. But to a machine trained on past patterns, they can look that way.

There’s something even more troubling. These algorithms learn from historical data, which is almost always imbalanced. Because fraudulent transactions are rare on a per-transaction basis, the model has seen relatively few examples of what fraud looks like across every type of customer.

What does this mean? Research has found that customers in lower-income areas and communities of color face higher rates of erroneous declines. When a model hasn’t seen enough transactions from a particular group of people or in a given situation, it has less data to build an accurate baseline for them. So when something slightly unusual happens, it flags it. Not out of intent, but out of unfamiliarity.

The model isn’t necessarily explicitly discriminating against anyone. But its outputs can still produce what researchers call disparate impact – unequal harm, distributed unequally.

As researchers at MIT explain in their book “Fairness and Machine Learning,” this is a known limitation. A model trained on incomplete representation will perform less reliably for the groups it saw least. The fix isn’t to blame the algorithm, but to train it on better, more representative data, and to test its error rates across different customer groups before deployment.

Why you don’t have the right to an explanation

What makes these cases worse is the lack of any information.

When a loan officer denies your mortgage application, the law requires a written explanation. But when an algorithm declines your debit card, you get “flagged by our system” message. If you’re lucky enough to connect with a human representative, they can’t tell you much more.

This gap isn’t an accident. Most high-performing fraud models are black boxes. Their internal logic isn’t designed for human interpretation. A bank may genuinely be unable to articulate plainly why your transaction was stopped. That’s not because it’s hiding something, but because the model itself doesn’t produce a reason. It produces a number.

In response, some financial institutions are moving toward tools that make their algorithms more transparent. Known in the industry as “explainable AI,” these systems are designed to surface the most influential factors behind a given decision – flagging, for instance, that a transaction was blocked because of an unusual location combined with an atypically large amount. It’s a meaningful step toward accountability.

However, these adoptions are uneven, and the explanations that do exist are rarely surfaced to customers.

Meanwhile, those pressures haven’t yet translated into a consistent, enforceable right to a meaningful explanation when your card gets declined. Challenging a decision made by AI can be enormously difficult, and most of us don’t even know we have the right to try.

For most people, the path of least resistance is simply to move on, switch to another card, take their business elsewhere or say nothing. Research suggests a quarter of consumers who experience a false decline never return to that merchant at all.

Some people go further and close the account entirely. That instinct is understandable. However, it carries a hidden cost. A declined transaction won’t appear on your credit report, but closing the card can. Shutting down an account reduces your available credit and can shorten your credit history, which can directly affect your credit score.

What you can actually do right now?

You have more power here than the banks would like you to think.

Call your bank immediately: A fraud flag is probabilistic, not final. A bank representative can override a declined transaction in real time. The model made a guess, but a human can correct it. Do not wait.

Set alerts if you’re planning to make unusual purchases: Most banks allow you to notify them of upcoming travel, large purchases, or changes in your spending pattern. This doesn’t override the model, but it gives it new information to work with, which can prevent the flag from triggering in the first place.

Know your rights: Under the Fair Credit Billing Act, you can dispute erroneous transaction blocks and request an explanation. If you believe you’ve been systematically and unfairly blocked, the Consumer Financial Protection Bureau accepts consumer complaints.

Ask your bank what appeal processes are available: Increasingly, banks are building more customer-facing appeal services. Visa reported 106 million disputes globally in 2025, a 35% rise since 2019, and has called dispute management a “strategic priority.” Improper declines are expensive for payment companies and financial institutions, too, through customer service costs, lost revenue and eroded trust.

The bigger picture

The algorithm that blocked your payment isn’t all-knowing or neutral. It’s a machine making a statistical guess about you, based on data that was probably never perfectly fair to begin with.

As AI spreads further into our daily lives, the question of who controls these decisions, and whether we can challenge them, becomes ever more urgent. The technology will keep expanding into new realms. The rules, and our own financial fluency, need to keep up.

Pragati Awasthi does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

On Feb. 10, 2026, an 18-year-old woman, Jesse Van Rootselaar, killed eight people and herself in a mass shooting in Tumbler Ridge, British Columbia. OpenAI had previously flagged her ChatGPT conversations as having a disturbing fascination with extreme violence, and suspended her account, but reportedly the company did not notify law enforcement.

On Oct. 2, 2025, a young man named Jonathan Gavalas in Jupiter, Florida, took his own life after developing what his father’s lawsuit described as a romantic attachment to Google’s Gemini chatbot. The suit claimed that Gemini coached Gavalas to shed his own body. The suit said Google had flagged Gavalas’s account 38 times over five weeks for sensitive content, but didn’t restrict or cut off the account.

These tragedies and others show that generative AI can potentially play a role in harming people, organizations and the environment. I’m a legal scholar who has focused on AI liability for nearly a decade and explored new ways of analyzing AI companies’ responsibilities. In my view, cases like these force questions the legal community has not come to terms with: If an AI company becomes aware of warning signs about harm, does it have a legal obligation to at least warn the appropriate authorities? And if the company doesn’t intervene, should its failure to act be considered negligence?

A need to raise red flags

U.S. tort law provides a framework for thinking about this type of responsibility. In 1969 a University of California psychiatric patient named Prosenjit Poddar told his therapist he intended to kill a woman named Tatiana Tarasoff. The therapist notified campus police, who briefly detained Poddar but eventually let him go. Nobody warned Tarasoff, and Poddar killed her shortly after.

Her family sued the university, arguing that its lack of warning amounted to negligence. In 1976 the California Supreme Court ruled that when a mental health professional has good reason to believe a client poses a serious danger to an identifiable person, they have a legal duty to take reasonable steps to protect that person, including warning them or notifying law enforcement. Today, most U.S. states recognize some version of the Tarasoff duty to protect or warn.

The logic is simple: If you have special knowledge of a serious threat and are in a position to address it, even if only to warn the authorities or the potential victim, the law may require you to act. But does that logic apply to AI companies?

The argument for yes is appealing. AI platforms interact with millions of users daily, often about deeply personal matters such as mental health struggles, relationship problems and violent thoughts. Most companies have systems to detect conversations that raise red flags.

Requiring a response might be less controversial for AI than for a human therapist. Therapists are bound by strict confidentiality obligations that make warning third parties ethically and legally complicated. AI companies operate under much weaker rules, at least in the U.S., where no comprehensive federal privacy law exists.

That lesser restriction makes it easier to justify requiring AI companies to act when it seems that someone’s life may be at risk. But balancing that with protecting privacy is still important.

Who to warn, and when

The first challenge in applying the Tarasoff framework to the AI world is accuracy. Predicting violence is hard, even for trained mental health professionals. AI systems, or human moderators who review flagged content, are not clinicians. Requiring them to judge who poses a genuine threat could lead to numerous false positives, with real consequences for people whose accounts are suspended or whose information is shared with authorities based on misread signals.

The second challenge is scale. A therapist sees dozens of patients. AI platforms have hundreds of millions of users. Imposing a duty to monitor and act on worrisome content could create perverse incentives. AI companies might reduce their monitoring to avoid acquiring knowledge that would trigger a legal duty, reasoning that what they do not know cannot make them liable.

The third challenge is identifying who is at risk. In the 1969 case, Poddar had named Tarasoff as a potential victim. But in many AI interactions, violent or self-destructive language is diffuse and doesn’t identify a target. Courts will need to develop clear standards for when a threat is specific enough to trigger a duty to warn, and to whom any warning or protective action should be directed.

Growing urgency

The AI industry is expanding rapidly, yet the legal rules governing what AI companies owe their users and the public are deeply unclear. Courts are beginning to grapple with questions case by case, such as whether OpenAI bears any responsibility for a gunman accused of killing two students at Florida State University on April 17, 2025. The gunman in that case was armed with a semi-automatic pistol and allegedly had extensive conversations with ChatGPT about how to use the weapon most effectively .

A narrow, carefully defined duty to warn, triggered only when an AI system flags a user’s behavior and it is reviewed by humans, would be a meaningful step forward. And it could focus initially on the most serious and credible threats.

The practice could also shift the conversation away from thorny technical debates about whether AI chatbots are products, services or media, which complicates legal claims, toward a more human question: Did this company know someone was in danger, and did it do enough to warn them and authorities?

Anat Lior is affiliated with: 1. Mentoring at the Creative Destruction Lab (CDL) at the University of Wisconsin 2. Member of the Montgomery County Advisory Council on Artificial Intelligence for the Public Good/ 3. Consultant with WTW, Relm Insurance, and Testudo. 4. Affiliate research with the Institute of Law & AI. 5. Collaborator with the Vista Institute for AI Policy.

Car crash. Cancer diagnosis. Mental health crisis. Autoimmune disease flareup.

A serious medical condition can turn your life upside down in an instant, making everyday tasks feel overwhelming. And if you’re employed, you may find that work emails keep coming and your manager keeps calling – when the only job you should focus on is healing.

In these moments, a medical leave of absence from work can serve as a vital lifeline.

We are organizational behavior professors who research how people balance their personal and work lives. In a study published in the Journal of Applied Psychology in March 2026, we found that employers can design and enact medical leave policies in ways that support healing rather than adding more stress during what is already one of the hardest periods of an employee’s life.

We conducted interviews with 30 employees who had taken medical leave from a wide range of professions, such as teaching, management consulting, nursing and landscaping. We also interviewed 18 human resources professionals who manage the medical leave process. By systematically analyzing what people said during the interviews and looking for patterns, we determined what many employers are doing to help their workers heal.

2 in 3 Americans can take paid medical leave

Employees take medical leave when sick leave is not enough – when recovery will require weeks or months off.

But many workers make their jobs a higher priority than their health. Some fear being seen as less committed or losing their job if they take leave. Others simply cannot afford to lose income. As a result, many people work while getting chemotherapy, postpone surgeries doctors have told them they need, or forego other necessary treatments altogether, even when laws and workplace benefits may exist to protect them.

About 2 in 3 employed Americans had access to paid leave for their own serious health condition as of 2022, and about 9% of the people who had paid leave didn’t use it when they needed it.

Under the Family and Medical Leave Act, eligible U.S. workers who have worked for a company with more than 50 employees for at least one year can take up to 12 weeks off to heal from their own serious health condition, or to care for a loved one such as a new baby or seriously ill family member.

But that policy protects your job, not your paycheck. It’s up to your employer, or your state, to determine whether medical leave is paid or unpaid.

Colorado, Connecticut, Massachusetts and several other states mandate paid medical leave for their employed residents. Some big employers also provide paid medical leave, including Microsoft and Adobe.

What to do when you need it

If your symptoms or treatments are making it hard to do your job, don’t wait to get started. Chances are that you need to take time off from work to heal. And you should not delay treatment to accommodate what’s going on at work.

Experiencing stress from your job when you’re ill or injured can be like gasoline on a fire – it can exacerbate health problems and make it much harder to bounce back. We were surprised by how many people we interviewed waited until their circumstances were dire before stepping away from work.

It’s also important to check what benefits are available to you.

You may qualify for protections under the Family and Medical Leave Act, which can keep your job safe while you recover. But it doesn’t apply in all cases, such as when employers have fewer than 50 workers.

To protect your paycheck, you may have access to a short-term disability policy through your employer benefits package that you can use in conjunction with the Family and Medical Leave Act.

Alternatively, you may already be paying into your state’s paid leave program through payroll deductions. These programs work like insurance, helping replace part of your income while you are on leave.

Your human resources department can serve as your first point of contact and can get you in touch with a leave coordinator, if your employer has one.

However, you do not have to share detailed information about your medical condition with your supervisor, or even HR, if you prefer to keep that information private. Your doctor only needs to provide documentation confirming that you have a serious condition and detailing how much time you need off.

Some employers also offer extra support through employee assistance programs, which can provide free counseling sessions, or financial and legal assistance.

Some best practices

We found that access to paid leave is important, but not sufficient, for helping workers heal.

Many large employers that effectively support workers in need of medical leave have trained specialists in their human resources departments who help employees understand their options. That makes it easier for workers to take enough time off to recover.

Employers that handle medical leave well also train managers on the basics.

They make sure managers know how to clearly communicate the available benefits to their subordinates, understand who is eligible for them, and know who from human resources can support workers throughout the process.

But a manager’s role ends there. Managers do not have discretion over when or whether an employee may take leave. Good managers know that, and understand that their role is to support their employees during what is likely one of the most difficult moments of their life.

Employers can proactively prepare for workers to take extended absences by cross-training employees ahead of time. Doing this signals that taking leave is acceptable, expected and supported. If the need for leave arises, workers are less likely to feel guilty about stepping away to focus on their health because they know someone else can temporarily cover their work.

We believe that the best employers ensure medical leave benefits are available from day one on the job.

Under federal law, workers must be employed for at least 12 months before they qualify for the Family and Medical Leave Act’s protections. But illness and injury do not happen on convenient schedules.

A car accident after 11 months on the job is just as devastating as one after 11 years. Someone who starts having unexpected seizures eight months into the job still needs the time away from work to seek treatment and a diagnosis. Employers that really want to support their employees and their well-being recognize this and do not make employees wait.

Even if you’re not a manager, you can play a role. If any of your coworkers are getting ready to take medical leave or are already on leave, you can support them by learning more about their daily tasks and helping fill the gaps.

Liza Barnes receives funding from the Society for Human Resource Management and the Academy of Management.

Ashley Hardin and Christina Lacerenza do not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.

Physical activity is one of the most powerful health tools people have to improve mood, energy and sleep, even after just a few sessions.

But the real superpower of an active lifestyle is what it can do for health and quality of life over time. Scientific evidence repeatedly demonstrates that physical activity reduces the risk of developing chronic conditions such as heart disease, diabetes and even some cancers. Despite this, most Americans are not getting enough physical activity in their daily lives.

So why are so few people physically active when the benefits are widely known?

As a physical therapist and rehabilitation scientist who studies how to boost movement for people living with chronic conditions and physical disabilities, I spend a lot of time thinking about that question.

The short answer is that understanding the importance of exercise usually doesn’t translate into exercising. Making it a part of your lifestyle requires believing you can do it and knowing you can do it.

Exercise is a lifestyle choice that helps reduce the likelihood of developing a chronic illness. But the good news is that if you’re one of the 194 million Americans already living with one or more chronic illnesses, beginning or maintaining an exercise routine can slow the progression, reduce symptoms and improve health outcomes.

An array of health benefits

The list of benefits from movement is long. Here are just a few examples:

• Physical activity improves the heart’s efficiency to move blood to the lungs and circulate it throughout the entire body.

• Engaging in physical activity protects against cognitive decline over time.

• Becoming more active contributes to improved insulin resistance, which protects against diabetes risk and helps manage diabetes symptoms.

• An active lifestyle can reduce cancer risk and improve outcomes for various cancer types, including breast, colorectal and prostate cancer.

Guidelines for getting and staying active

While some movement is better than none, the Centers for Disease Control and Prevention offer research-based guidelines for the type and frequency of activities to engage in weekly for long-term health benefits.

The CDC encourages all adults, including those with chronic health conditions or disabilities, to aim for at least 150 minutes of moderate intensity aerobic activity, such as walking, jogging or swimming. Adults should also do muscle-strengthening activities two or more days per week, which could include weightlifting and body weight exercises.

Older adults should add balance activities, such as tai chi or yoga, to help prevent falls by challenging the body’s balance systems.

If you’re not achieving these recommended weekly physical activity guidelines, you’re not alone. Only half of Americans hit the aerobic target, and just 1 in 4 meet the full CDC guidelines.

This gap represents a health crisis that, if addressed, could save lives. A 2024 large-scale review showed that people who engage in regular physical activity in adulthood may reduce the risk of early death by 30% to 40% from all causes, most specifically from cardiovascular disease and cancer.

The study also showed that beginning exercise at any time in adulthood can improve survival benefits.

The difference between knowing and doing

People are perpetually being sold on the benefits of physical activity, whether it’s from national healthcare organizations, their medical teams or social media influencers.

But research is clear that education alone does not predict changes in behavior.

Instead, shifting your beliefs about the barriers preventing you from exercise might actually be the key to get you moving more.

In 1977, a psychologist named Albert Bandura proposed that the ability to perform a task even when it’s difficult – a concept called self-efficacy – is the most important personal characteristic that drives healthy changes in behavior.

Half a century later, self-efficacy is still considered one of the most crucial personal factors for behavioral change when it comes to long-term physical activity. Researchers who develop and test exercise interventions, including me, evaluate novel tools and programs that are built to boost self-efficacy.

Someone with high self-efficacy might say that they can get back to their exercise routine even if they miss a day. Or they might find a way to still exercise when they’re busy or tired. Someone with lower self-efficacy might be thrown off their routine if presented with the same obstacles.

But how do you build this crucial trait and get moving more? A meta-analysis found that despite its importance, there is not one magic way to boost self-efficacy.

That’s because people’s behavior is more complicated than individual factors alone. People and groups have varying needs and contexts that require tailored approaches.

Tips increase exercise self-efficacy

Self-efficacy may be affected by multiple factors, but people can still apply techniques to boost their ability to start and stay with an exercise routine.

Make it manageable. It may seem intuitive to set personal goals, but many of us aim too high and end up discouraged. Goals focused on weight loss, heart health or muscle strength are fine, but they can take a long time to achieve. Long-range goals don’t tend to be motivating in the difficult moments – like when you want to hit snooze but promised yourself that you were going to take a long walk before work.

Instead, try short-term goal-setting – such as aiming to get a set number of lunchtime walks in during the workweek. This will move you toward your long-term goals, while making it easier to see and feel progress.

In 2026, the American College of Sports Medicine refreshed its guidance on strength training, which represents synthesized findings from 137 systematic reviews and the first update since 2009. The biggest recommendation difference? Consistency matters more than specificity of strength programs. What that means is that doing any strength training has health benefits as long as it is the kind you will keep doing.

Make it add up. The CDC’s recommended 150 minutes of aerobic activity is meant to be spread throughout the week – not done all at once. Research shows that small bursts of activity still have significant impacts on your overall health, and you’re much more likely to stick with them.

Only have 15 minutes while your kid is asleep? Have a short exercise video or app cued up for nap time. Waiting for your next Zoom meeting to start? Climb your stairs once or twice. Microwaving your lunch? Hold on to the counter and lift and lower your heels until the timer goes off. Every little bit matters to your mind and body.

Make it meaningful. Prioritize doing things you enjoy. The gym is not for everyone, and luckily this style of structured exercise is just one of many options for physical activity. Go bird-watching, join a gardening group, binge watch your favorite show on the treadmill. Any activity you do that uses energy is like dropping a coin into your weekly physical activity bank.

Make it more fun. Choose to be around people who are already exercising – and who encourage you to do it, too. Research shows that people who are sedentary will increase their physical activity by socializing with someone who is active.

Another study shows that older adults can tap into the energy of their peers during group exercise, helping to build self-efficacy. Exercising with others can even reduce social isolation and loneliness. As a bonus, choosing physical activities you enjoy can improve your mood and boost your confidence.

Overcoming the hurdles

These strategies come with a very important caveat: Increasing self-efficacy is empowering, but context also matters.

Some structural barriers to physical activity are beyond the scope of our individual motivation. Researchers and health professionals know that lower socioeconomic status, decreased neighborhood safety and lack of access to exercise programs make being and staying active even more difficult.

But the thing to remember is that even small improvements can have big impacts. It is consistent practice – not perfection – that is key to reaping all the benefits physical activity has to offer.

Laura Baehr receives funding from the Department of Defense, the Arthritis Foundation, and the Clinician-Scientists Transdisciplinary Aging Research Coordinating Center (a National Institutes of Health National Institute on Aging funded center).

My mother loves butter. It is the primary fat I ate growing up. She smeared it on any kind of bread, potatoes, nut rolls or coffeecake. She baked with it exclusively.

When I was studying nutrition in college, I had a teaching assistant who recommended margarine over butter. I was shocked – and wondered about the difference between the two. It was one of the things that sparked my interest in food science. Today, I am a food scientist and study how foods such as butter and margarine can have subtle chemical differences, with a big impact on how they act in food.

Chemical structures

Butter and margarine are emulsions, which are mixtures of tiny water droplets spread throughout a continuous fat matrix. This matrix is made mostly of triglycerides, the primary form of fat in our diet.

Fatty acids are long chains of carbon surrounded by hydrogen atoms. In a triglyceride, there are three fatty acids, each one connected to the same three-carbon glycerol molecule, which acts as the backbone of the molecule. While the backbone is always the same, the number of carbons in the fatty acids can vary. In cream, triglycerides are packaged into globules or crystals.

Both butter and margarine have a combination of saturated and unsaturated fatty acids. However, butter’s are mainly saturated, which makes them fit together and stack compactly to form a nice straight chain, because they have no double bonds between the carbons.

Margarine’s fatty acids are mainly unsaturated, from blended plant oils. The unsaturated fats give them an irregular shape on a molecular level. The double bonds between carbons kink the molecule so they cannot be as neatly arranged. This difference affects how they melt.

There are many forms of fat crystals in butter, and they have different melting points. These crystals make butter very firm at cold temperatures and allow it to soften gradually at room or body temperature. They also trap air easily when creamed together with crystalline sugar, which adds lightness and porosity to baked goods.

Both butter and margarine are at minimum 80% fat, though some butters are closer to 85% fat. Their water content hovers around 16%, and butter is made up of 1-4% vitamins, minerals, lactose and protein.

Butter has an official standard of identity set forth by the U.S. government, which means manufacturers must meet specific guidelines for their product to be considered butter. This food standard is one of the oldest in the U.S.

Making butter

When you shake or churn cream, fat globules rupture. The fat leaks out and forms semi-solid grains of butter. With more shaking or churning, these grains grow and separate from the watery, naturally low-fat buttermilk.

You then collect, knead and press the mass and, voila, you have butter. Some butter is cultured by adding lactic acid bacteria. These bacteria ferment milk sugar, or lactose, into flavor compounds and organic acids, which give the butter a mild tang and complex flavor.

Sweet butter is easy to make at home if you add cold, heavy cream with a fat content of at least 36% to a standing mixer with a whisk attachment. Turn it on and walk away for a bit, and when you hear the sloshing sound of watery buttermilk, you know you have butter ready for pressing.

Making margarine

Sticks of margarine start as liquid, plant-based oils and are made into a solid. Producers chemically rearrange fatty acids on the glycerol molecule in a modification process called interesterification, which makes the oil solid and the fats more uniformly distributed.

This process rearranges the triglycerides in margarine without adding saturated fats or creating trans fats. Trans fats have been banned in many countries because of their association with cardiovascular disease and higher cholesterol.

Interesterification allows margarine to stay solid longer when baking, with a more precise melting point.

Spreads or squeeze-style margarines do not go through this process and instead rely on higher ratios of water and air to solid oils, which keep them soft and spreadable. These spreadable types are lower in fat, so they don’t work well for baking. The higher water content alters the texture, and most baking recipes are formulated assuming a higher percentage of fat.

Processors are not required to state on the label whether margarine has gone through interesterification.

Flavor and color

Butter gets its golden color from beta-carotene, an orange pigment present in grass. Cows eat the grass but do not metabolize beta-carotene efficiently, so it is expressed in their milk. Margarine is naturally colorless, but producers add synthetic beta-carotene to it to mimic the color of butter.

Margarine producers also add flavors such as diacetyl, a distinctive butter-flavor molecule, and blends of whey components and preservatives to replicate the flavor of butter. They may add emulsifiers such as lecithin or monoglycerides to keep the water and fat from separating. The exact ratios of ingredients vary between producers.

Chemical differences can translate into subtle health differences. While both are mainly made of triglycerides, the fats in butter are naturally occurring, while fats in margarine are industrially modified. This difference makes margarine an ultraprocessed food, but it also means it has fewer saturated fats. While you might have health reasons for choosing one over another, take note that the chemistry behind how these fats are made also can influence how they behave in the kitchen.

Baking differences

When you heat butter, the proteins and lactose in it combine, creating that signature brown color and a delicious nutty, toasty, caramelized flavor. Because margarine doesn’t contain lactose, it won’t brown as well as butter, nor will it impart the same level of aromatics.

When baked in a very hot oven, butter contains enough water to form steam, which separates doughs into layers of flaky pastry. Water content varies in margarine, and while it forms some steam, it will not perform as well as butter.

However, margarine has some advantages over butter. It’s very consistent and melts in controlled way. It also has a longer shelf life. Yes, you can use them interchangeably, but knowing functional differences between the two can help you determine when to use which.

Rosemary Trout does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.