Manipulation Defense curriculum

Philosophers usually set manipulation apart from rational persuasion, which works by offering reasons. One family of accounts says manipulation bypasses deliberation; another says it uses nonrational means that leave deliberation worse off. Neither is settled. Read literally, bypassing misses many genuine cases; read loosely, it sweeps in influences, such as some emotional appeals, that do not seem manipulative.

Following Joseph Raz, some philosophers say coercion interferes with a person's options, while manipulation leaves them intact and corrupts the way decisions and preferences are formed. Pressure-based accounts picture a scale instead: persuasion applies no pressure, coercion applies the most, and manipulation sits between, pressing in ways that stop short of coercion.

Covert-influence accounts, including one by Susser, Roessler and Nissenbaum, define manipulation as influence the target is unaware of and could not easily notice. That fits framing effects, misdirection and gaslighting. It struggles with peer pressure, nagging and guilt trips, which are hard to miss, and even a version requiring only a hidden motive runs into trouble when the motive is plain, as with a teenager nagging for a phone.

Trickery accounts tie manipulation to deception: the target is led into a false belief, an unwarranted suspicion or an out-of-place emotion, as Iago does to Othello. Pressure accounts center on nagging, peer pressure and emotional blackmail, which need not deceive anyone but attach costs to saying no. Real manipulators often mix both, and whether one definition can cover the two is still debated.

Medical ethicists widely hold that manipulation undermines the validity of consent, and proposed conditions for informed consent often require that agreement was not manipulated. The hard part is the line. Patients may choose an operation more often when told of a 90% survival rate than a 10% fatality rate, yet some framing is unavoidable, since one figure has to come first, and no consensus says which nudges cross over.

A scoping review of 50 studies on romance scams describes how scammers may comb dating profiles and social media, then use what they find to tailor their persona and approach to the target. Government impostors may already know a name, address or even a Social Security number, which helps a fake call sound real. Personal details in a caller's hands therefore prove nothing about who is calling.

If every influence besides rational persuasion or coercion counted as wrongful manipulation, nothing but rational persuasion would ever be legitimate, a conclusion few accept. A related caution, prompted by Sarah Skwire's critique of a book that called many sales practices manipulative, is that quickly suspecting manipulation whenever people choose differently can dismiss choices the critic does not understand.

The elaboration likelihood and heuristic-systematic models describe a central route, in which people weigh a message's substance, and a peripheral route, in which cues about its source carry the decision. Both hold that central processing judges argument strength more accurately. Cues are not always misleading: in one mock-jury study, jurors arguing for the sound verdict swayed their groups more when they also raised cues such as credentials.

Dual-process models hold that careful scrutiny needs both the motivation and the ability to apply it. In a lab study of 224 university students and staff, those told to judge emails within five minutes told phishing from real messages slightly less accurately than those working at their own pace. In one online scenario study, heavy cognitive load strengthened the link between negative feelings and sticking with a failing investment.

In a classic copy-machine study by Langer, Blank and Chanowitz, a request whose reason added nothing, that the person needed to make copies, drew about as much compliance as one with a real reason, and both beat asking with no reason. The effect appeared for a small favor, not a larger one. One direct replication found the main effect, but later conceptual replications found smaller effects.

Need for cognition is a person's intrinsic motivation to process information thoroughly. Earlier research suggests high scorers lean on substance, while low scorers are more swayed by cues such as an expert's credentials or fee. In one mock-trial study of college students, the gap showed in just one condition: when cross-examination raised only such cues, discussion left low scorers marginally less accurate while high scorers held steady.

In field experiments on the emotional seesaw, people jolted from fear to relief, or from delight to disappointment, agreed more often to requests, from minding a stranger's luggage to donating money. Two lab experiments with 80 and then 41 participants found such swings raised signing of a nonsense petition and reduced message processing and recall. The evidence concerns abrupt shifts between emotions, not strong emotion in general.

A systematic review of research on older adults found no evidence that they make up a larger share of online fraud victims than other age groups, and no consensus yet on how cognitive function or health shapes their risk. In a phishing study of students and staff, cognitive reflection and sensation seeking predicted errors only modestly, while merely being rushed also raised errors. Risk appears to depend on the situation as well as the person.

Psychological inoculation pairs a warning with a weakened dose of a misleading technique and a rebuttal of it. Across six preregistered experiments and a YouTube field study, short videos on tactics such as scapegoating and false dichotomies improved people's recognition of them. But in five later studies of a simulated feed, an inoculation against emotional language cut engagement only when no real posts or other tactics were mixed in.

Because the reciprocity norm says favors should be returned, requesters can use it to their advantage by doing a small, unexpected favor before asking. Romance scammers also often send gifts, or promise them, while building trust. US law draws a firm line for unordered goods: you never have to pay for merchandise you did not order, even when a charity sends it with a request for a donation.

The tactic opens with a large, objectionable request and, after a refusal, retreats to a smaller and more reasonable one, which is said to raise the odds of agreement. Researchers explain it through reciprocity: the retreat is presented as a favor, and a favor creates a sense of obligation to return it.

The tactic sweetens an offer by announcing a special bonus. A meta-analysis of 18 effect estimates found a modest average advantage, a correlation of about 0.16, but its range of likely effects in a new setting included negative values, so the tactic does not reliably help. A wider review of 30 message tactics found most make small, inconsistent differences in persuasiveness.

Tech support scams often start with a bogus pop-up warning, styled to look like it comes from a well-known company, that urges a call to a phone number. The scammer then asks for remote access, pretends to scan, claims to find malware and offers to remove it for a fee. The FTC notes that real security warnings never ask you to call a phone number.

In an analysis of 2013 Medicare prescribing data in Washington, DC, physicians who accepted drug-company gifts worth under $500 a year wrote more expensive and more often brand-name prescriptions than those who took none, and larger gifts went with larger differences. The data are observational, so they show an association, not proof that the gifts caused the change.

When US securities regulators examined 110 firms and branch offices offering free lunch investment seminars, every seminar reviewed turned out to be a sales presentation, many advertised as educational, and half made exaggerated or misleading claims. The FTC likewise describes investment scams that start with free training or seminars and quickly lead to hefty fees for more coaching.

Federal law bars companies from sending unordered goods and then demanding payment, and recipients need not return them. Goods clearly marked as gifts or samples, and goods a charity sends along with a donation request, may also be kept as free gifts. In law, at least, an unasked-for gift of goods creates no debt.

Starting with a small request and then making a larger one raises compliance on average, meta-analyses suggest, and commitment theory explains it as people staying consistent with what they have already done. Romance scammers commonly use the same ladder, asking for small sums to test willingness before larger ones. Yet the research also contains many failures to replicate and some reversals.

Researchers studying car salespeople found a tactic the salespeople called low-balling: once buyers have decided on a bargain car, they are more likely to go through with a purchase even after the price rises or that car proves unavailable. An FTC complaint alleged a related yo-yo financing scheme, in which dealers told buyers who had driven off that financing fell through and pressed them to sign worse terms.

The FTC treats it as illegal to advertise a product a company does not intend to sell, planning instead to sell something else, usually at a higher price. A 2022 FTC proposal for car dealers targeted bait claims about price, financing, discounts and whether advertised vehicles were actually available, noting that once buyers are in the door, they face the fallout of promises that do not pan out.

Commitments others can see have been linked to more recycling, towel reuse and voting. In randomized trials during the 2016 election, young people who pledged to vote turned out about 3.7 points more than those sent standard get-out-the-vote materials. But in a trial across 196 English general practice units, prescriber commitment posters promoting careful antibiotic use did not reduce prescribing.

In Aronson and Mills's classic study, women who went through a more embarrassing initiation to join a discussion group rated it more favorably than women given a milder one. Dissonance theory reads this as easing the clash between having suffered and preferring not to. A 2020 EEG study of 82 young adults found no low- versus high-effort difference in a brain reward signal, though within the hard task, more felt effort went with a larger signal.

Past investments cannot be recovered, so economic theory says they should not steer the next choice, yet people often keep investing in failing plans. Scenario studies link this to the bad feeling of wasting what was spent. A large preregistered online replication found sunk-cost effects for time as well as money, though the original claim that time pulls less than money held in only one of its two tests.

Economic theory holds that only future gains and losses should guide a decision. In one online scenario study, people asked to justify their choice were more resistant to the sunk-cost fallacy and felt less negative about the options. And when a car dealer cancels a deal because financing fell through, an FTC account notes that the dealer generally must return the down payment and trade-in.

The 2024 FTC rule bars reviews by people who do not exist or never used the product, including AI-generated ones, and bars paying for reviews that must express a set sentiment. It also limits undisclosed insider reviews and company-run sites posing as independent. In one earlier FTC case, a dealership's internet manager allegedly posed as a customer to post a five-star review.

Under FTC guidance, an influencer with a material connection to a brand, such as payment, free products, or a family or work relationship, should make that relationship obvious in the endorsement. The stated reason is that disclosure lets people weigh the value of the recommendation. A platform's built-in disclosure tool is no guarantee that the disclosure is clear enough.

An ad citing an endorser's specific results is usually read as what others can expect, and the FTC says labels such as results not typical do not change that reading. Advertisers must either prove the results are typical or clearly disclose what people generally achieve. The FTC also warns that success stories in investment-training pitches are often made up or a rare exception.

In Goldstein and colleagues' hotel study, a card saying most guests who had stayed in the same room reused their towels produced the highest reuse, about 49% against 37% for a standard message. Replications have been mixed: a 2023 preregistered test of the same message found no clear effect on stated reuse intentions, which were already high, and a meta-analysis found descriptive norms have a small overall effect.

The FTC's 2024 rule prohibits using groundless legal threats, intimidation or certain false public accusations to block or remove a negative review. It also bars implying that the reviews shown are all or most of those submitted when low-rated ones were held back. A page of praise can therefore reflect filtering rather than customer satisfaction.

The FTC's 2024 rule bars knowingly buying or selling fake indicators of social media influence, such as followers or views generated by bots or hijacked accounts, when used for commercial purposes. FTC consumer advice adds that a sudden burst of reviews, or a reviewer account created to post a single review, can be a sign of fakes.

The FTC's long-standing advice is to consult a variety of sources, including well-known sites with trustworthy, impartial expert reviews. It also cautions that fakes can be nearly impossible to spot by eye: some fake positive reviews give less than top marks to look believable, and some fake reviews are negative, posted to harm a competitor.

Affinity fraudsters are often members of the group or pretend to be. Many affinity schemes are Ponzi or pyramid schemes, paying earlier investors with new investors' money so the venture looks successful. The tight-knit structure of many groups can also make these schemes hard for regulators and law enforcement to detect.

Securities regulators describe fraudsters convincing respected community or religious leaders that an investment is legitimate and relying on them to spread the word. Many of those leaders become unwitting victims. Regulators warn that the person recommending an investment may have been fooled too, and that victims are especially likely to keep problems inside the group when respected leaders drew them in.

Affinity fraudsters are often, or pretend to be, members of the group they target. A scoping review of romance scam research describes scammers who may study a target's profiles and social media and who, once in contact, stress similarities by adapting their story to what the victim shares. Marketing research likewise lists similarity among persuasive wording tricks, since people tend to like those who resemble them.

Research on romance scams describes a grooming phase of intensive communication, frequent compliments, emphasized shared interests, declared feelings and sometimes gifts. It can last months or even years, with regular contact routines keeping the target engaged. Researchers describe the aim as building enough trust and emotional commitment to leave the target more open to later requests for money.

The FTC describes romance scammers who soon move contact off the dating platform, say it is true love, and explain that they live far away and cannot meet. Researchers also describe love bombing, a burst of intense, repeated affection in a short period, which can emotionally disarm a target and shrink the time and space for careful financial decisions.

Securities regulators note that affinity fraud victims often fail to notify authorities or pursue legal remedies, trying instead to work things out within the group, especially when respected leaders drew them in. They also list being told to keep an opportunity confidential, or that there is no time to put it in writing, as warning signs.

SEC guidance says to check out everything no matter how trustworthy the person bringing an opportunity seems, and never to invest based solely on a fellow group member's recommendation. The SEC's investor website offers a free search showing whether an investment professional is licensed and registered, and whether there are complaints or disciplinary actions.

Impostors claim to be from the IRS or Social Security, say they are calling about Medicare, or pose as a utility, then threaten arrest, lost benefits or shutoff unless payment comes at once. The FTC says government agencies never call, email or text to ask for money, and real utilities do not demand immediate payment to avoid same-day disconnection. Insisting on gift cards, wires, crypto or payment apps is a scammer's tell.

Scammers can make any name or number show up on caller ID, a practice called spoofing, even an agency's real number. Phishing emails can carry a real company's logo, and the FTC notes that top search results are often paid ads, which dishonest businesses sometimes use to pose as a utility. Each of these signals can be imitated.

Government impostors may give an employee ID number to sound official and may already know a person's name, address or Social Security number; fake FTC messages may even name real agency staff or commissioners. None of this shows the contact is real: the FTC says government agencies will never call, email or text to ask for money or personal information.

When a call or message might be real, the FTC advises hanging up and contacting the agency directly at a number known to be correct. It warns against calling back the number the caller gave or the one shown on caller ID, and points to the number on a bill or the official website, while noting that paid search ads sometimes pose as the real company.

A pooled analysis of 21 Milgram conditions with 740 participants found people rarely went to the maximum shock unless the experimenter explicitly directed it, and less often when peers urged disobedience or experimenters disagreed. Critics of the blind-obedience reading add that participants tended to continue when told the experiment required it, but typically refused the blunt prod that they had no other choice.

In one tech support scam variant described by the FTC, the fake technician claims the target's accounts were hacked or their name is linked to serious crimes, then transfers them to someone who supposedly works for the government. That person may supply a badge or case number, both fake, and says the money must be moved to protect it, which real government workers never say.

The FTC describes utility impostors as trying to scare people into paying before they have time to confirm what they are told, and notes that real utility companies do not do this. It also suggests talking it over with someone you trust when unsure, since that conversation can help reveal a scam. Government agencies, it adds, never call or text to ask for money.

Scam scripts start a clock so the target acts before thinking: pay today or face arrest, stay on the line so the story cannot be checked. FTC guidance treats the pressure itself as a warning sign and notes that honest businesses give people time to decide. Gift card scammers may even stay on the phone while the target goes to the store to load the cards.

The FTC's 2022 dark patterns report describes countdown timers meant to make shoppers believe an offer is time-limited when it is not. Studies of shopping sites cited in a 2026 review found many timers reset at zero with the offer still valid, and an automated scan of about 11,000 shopping sites counted urgency and scarcity among its dark pattern categories. A timer on the page is not evidence of a real deadline.

Scam scripts can present payment as the way to avert a loss: FTC guidance on gift card scams describes callers saying something terrible will happen unless the target pays right away. Research on the frame itself is deflating: a 2021 review that pooled 165 effect sizes from about 51,000 participants found almost no average difference in persuasiveness between gain-framed and loss-framed appeals.

In fake family emergencies, the FTC notes, the caller may stress secrecy because talking to other relatives would expose the lie. An FBI alert on business email fraud describes pressure to move money quickly or secretly, and romance scammers may ask that the relationship be hidden from friends and family. The FTC advises calling another relative or friend anyway, even when told to keep it secret.

FTC guidance describes calls claiming a grandchild is in trouble, an account has a problem or a computer has a virus, sometimes with a fake lawyer or police officer added for authority. The caller stresses urgency, says you are the only one who can help, and may know names and addresses gathered from social media, so familiar details are not proof of who is calling.

FTC guidance treats insistence on one of these methods as a core sign of a scam: no real business or agency demands payment by gift card, and a card's number and PIN let a scammer spend the balance even while you hold the card. Recovery is hard but not always hopeless: some gift card companies might give money back, and some can freeze a balance still on the card, so the FTC urges reporting right away.

FTC guidance pairs two steps: resist the pressure to act at once, because honest businesses give you time, and tell a friend, relative or neighbor what happened before doing anything, because talking it through can reveal the scam. For a call about a relative, it adds hanging up and calling back on a number you already know. Scammers counter by telling targets not to hang up.

The extended parallel process model of fear appeals holds that a threat's effect depends on how it interacts with efficacy, the sense that a recommended step will work and can be done. The scam threats the FTC describes come with that step already chosen: pay, in a specific way the caller names. The FTC's outside check: government agencies do not contact people to demand immediate payment, and never by gift card.

Research on guilt describes it as pushing people toward repair, such as an apology or help. A 2023 meta-analysis of 26 studies found guilt appeals had a small average persuasive effect, larger when the appeal named a way to make amends. Romance scammers may use a personal form, implying that a partner who truly loved them would do what is asked. The meta-analysis tested persuasive appeals, not relationships.

Sextortion means threatening to share intimate images unless the target pays or complies, and a 2024 scoping review notes that such images carry force because of the shame and stigma attached. Across its 24 studies, reporting and help-seeking stayed very low, with shame, fear and negative views of police and platforms among the reasons. Other scams seek a similar silence by telling targets to keep things secret.

The FDA says health fraud plays on the wish for a quick or miracle cure for hard problems such as cancer, diabetes or memory loss, some of it targets people with limited English or limited access to care, and it can delay proven treatment. The FTC describes immigration scammers promising special access to the visa lottery, or calling themselves notarios, a word that in Latin American countries means someone with legal training but not in the US.

A widely cited early account of trauma bonding lists three ingredients: a power imbalance, intermittent good and bad treatment, and captivity. Women interviewed after repeat domestic abuse described caring phases punctuated by cruelty, a cycle of unpredictability that deepened their attachment. That was 18 interviews, and the theory itself cannot explain attachment that forms without captivity or a power imbalance.

Researchers who build inoculation videos against misinformation list rhetoric meant to evoke outrage, anger or other strong emotions as one of five common techniques, with false dichotomies, scapegoating, incoherent arguments and personal attacks. All five come from research on argumentation and manipulation, which consistently flags them as unreliable ways to argue. The work concerns online posts and videos, not face-to-face arguments.

In six randomized studies with about 6,500 people, short videos that named techniques such as emotionally manipulative language improved recognition of those techniques and the ability to tell trustworthy from untrustworthy posts. In a YouTube field test the effect was small, and for the emotional-language video it missed the researchers' own preregistered threshold. FTC advice for emotional scam calls is simpler: slow down and verify.

FTC pricing guides describe the pattern: a seller briefly marks a product up, sells few or none, then advertises the usual price as a cut from the inflated one. In an experiment with 120 Taiwanese undergraduates, a higher stated original price for a free add-on raised what they said they would later pay for it, but only when that price seemed plausible. A former price need not have produced sales, so it may show only what was asked.

Adding an option that one choice beats on every count, while the other does not, tends to shift people toward the dominating choice. In online experiments, a decoy raised the share of previously uninterested people choosing cancer screening from 39% to 54% in one version. But across 3.6 million UK wine purchases, decoys moved preference by only about 1%, and an online experiment with about 90 tobacco users found no decoy effect.

A 2023 systematic review of consent to reuse health data found consent rates generally lower under opt-in than opt-out; in the one study that tried both, 21% opted in while 95.6% stayed in under opt-out. FTC guidance warns that a pre-checked box on a free trial can authorize later charges. The review cautions that its studies varied widely, so general conclusions are hard to draw.

In the classic disease problem, two groups saw identical programs described either as people saved or as people dying, and their risk preferences shifted with the wording alone. A risk-communication primer summarizes the usual pattern: gain frames often produce risk-averse choices and loss frames risk-seeking ones. In clinical settings, it adds, the effect varies from one situation to another.

The FTC counts this among dark patterns as burying junk fees. Its fee rule for live-event tickets and short-term lodging does not ban fees or itemized breakdowns; it requires the all-in total, with mandatory fees, to be the most prominent price. Outside those two industries, the FTC pursues drip pricing case by case under longstanding law against deception.

A risk-communication primer reports that when a baseline risk is small, a relative risk reduction is perceived as larger than the same absolute reduction, and that 1-in-x figures with varying denominators are harder to understand and raise perceived risk. Its remedy is to name the reference group and compare figures over the same denominator, such as 1 in 100 against 2 in 100.

Drip pricing shows a partial price first, and relative-risk figures can make a small change look large. The FTC's fee rule for tickets and lodging rests on the idea that an up-front total makes comparison shopping easier, and risk-communication guidance says figures sharing one denominator are easier to compare. The rule covers only those two industries; elsewhere the FTC acts case by case.

Sneaking hides or delays key information; obstruction adds needless steps; interface interference favors one choice through layout or defaults; forced action makes users do something extra, as with nagging pop-ups; social engineering leans on urgency, scarcity or social norms. A 2026 review used it to sort tricks on gambling sites but found almost no evidence yet on how they change behavior.

A notorious dark pattern, it covers accounts and subscriptions that take a few clicks to start but require hunting for a hidden option or writing to support to end. The FTC alleged that ABCmouse promised easy cancellation while routing users down a long, confusing path full of promotions. Gambling sites show a variant: accounts that are hard to close and easy to reopen.

The FTC's 2022 dark patterns report defines it as framing the alternative as a bad decision, as when a shopper declining a subscription must click a line saying they do not want to save money. A report on big tech privacy settings, summarized in a 2020 review, found choices worded so the protective option looked ethically questionable or risky, such as warning that keeping face recognition off left you open to impersonation.

The FTC's 2022 dark patterns report highlights ads built to look like independent editorial content and comparison sites that claim neutrality while ranking companies by what they pay. In one case, the FTC alleged that emails falsely claiming to come from news outlets led to fake news stories and then to a work-from-home sales pitch.

The FTC has sued companies for sneaking unwanted products into online carts without consumers' knowledge, and its 2022 report describes sellers that advertise part of a price and add other required charges late. A scan of about 11,000 shopping sites counted basket-sneaking among the dark patterns it found. In one FTC case, a lender allegedly promised no hidden fees while tucking fees behind tooltip buttons.

A well-documented banner design makes the accept-all button bright and prominent while shrinking or hiding the reject-all option, and a 2026 review notes that such designs have been found to sway acceptance substantially. An audit of UK consent platforms, summarized in a 2020 review, found only 11.8% met minimal legal requirements such as no pre-checked boxes and rejecting being as easy as accepting.

The FTC says a free trial's terms should state what you are agreeing to, how long the trial lasts and how to cancel; if you cannot find or understand them, the company may make canceling hard. It also flags pre-checked boxes that can authorize later charges or data sharing. Required fees can surface late in the buying process, so the first price shown may not be the total.

FTC guidance describes messages that appear to come from a bank, utility or payment app, claim suspicious activity or a billing problem, and ask you to click a link, open an attachment or confirm details. CISA adds lookalike sender addresses and links whose real destination differs from the text shown. A real logo proves nothing: the FTC's sample phishing email used the spoofed company's own logo.

CISA describes attackers who pose as a new employee, repair worker or researcher and use facts gathered from one person to sound credible to the next. In family-emergency scams, the FTC notes, callers may already know names and addresses from social media or a hacked email. The FBI warns that criminals use AI-written text in spear phishing to seem believable and to fix errors that might otherwise signal fraud.

CISA notes that internet calling makes caller ID easy to fake, exploiting misplaced trust in phone lines, and that tapping a link in a text can open a site or dial a number in one step. A live caller can also keep you on the line: the FTC describes scammers telling targets not to hang up, so the story cannot be checked. The FTC's counter is to hang up and call back on a number you already know.

FBI alerts describe a supplier asking for invoices to go to a new account, an executive's account instructing finance to wire funds, and people posing as lawyers on confidential, time-sensitive matters, sometimes timed to the end of the business week. In real estate deals, intruders monitor the transaction and time a request to change payment details. The FBI's counter is verifying any payment change on a separate, previously known channel.

A one-time code is the second lock on an account: a scammer who has your password still needs it. So callers posing as someone you can trust invent a problem with your account, offer to help fix it, and ask for the code. The FTC is categorical that anyone who asks for your account verification code is a scammer, and advises contacting your bank on a number from your statement or app.

The FTC notes that a short audio clip and a voice-cloning program can let a scammer sound like a family member, and the FBI warns of AI video used in live chats with fake executives or officials. The FBI lists flaws to watch for but also says AI content can be hard to identify; it advises agreeing on a secret family word and hanging up to call back on a number you look up.

FBI guidance on business email fraud recommends confirming payment changes by phone on a previously known number and setting up that second channel early, outside email. For family emergency calls, the FBI suggests a secret word or phrase agreed in advance, and the FTC suggests a question only the real person could answer. The common rule: never verify using contact details supplied by the request itself.

A recent crime-script review of 50 studies traced nine scenes, from building a fake profile and choosing a target to the payout and its aftermath. Scammers almost always try to shift talk to private apps, which also avoids platform monitoring, then build trust through constant contact, sometimes for months or years. The order is not fixed: when a target balks, scammers often return to grooming before asking again.

A review of romance-scam research describes scammers overwhelming targets with repeated declarations of affection, alongside talk of a shared future, gifts and requests to keep the relationship secret. Investor regulators note that romance scammers often pledge love very quickly. The evidence comes mostly from victims, investigators and court records rather than from scammers themselves.

Romance scammers may claim to be overseas, on an oil rig or in the military, and investor regulators note they may even suggest a meeting, then find excuses. The FTC's bottom line is never to send money or gifts to a sweetheart you have not met in person. A video call is not proof either: scammers may spoof calls, and the FBI warns of AI video made to prove a contact is real.

A review of romance-scam research describes early small requests, such as help with a plane ticket, that lower the barrier for larger ones, a pattern known as foot-in-the-door. Later stories grow in severity and cost, often medical, legal or travel crises, and victims are almost always promised repayment, which keeps many sending money despite doubts.

Regulators describe a long con: contact through apps, texts or a supposed wrong number, a slow build of trust, then coaching into crypto deposits on a site or app showing fake profits. Small early withdrawals may be allowed to build confidence. When the target tries to cash out, new taxes or fees appear, and the money is frequently never recovered.

A love interest or employer sends money, often by check or crypto, and asks that it be passed on by gift card or wire, or asks the target to open an account. The FTC warns that helping move stolen money can bring legal trouble even if you did not know it was stolen, and a fake check leaves you owing your bank. The FBI says many unwitting mules believe they are helping a romantic partner or employer.

The FTC calls it the bottom line: never send money or gifts to a sweetheart you have not met in person. It also suggests a reverse image search of the profile photo, searching the person's claimed job with the word scammer, and asking whether friends and family are worried. The FBI warns that criminals use AI to generate believable profile photos.

Every investment carries some risk, and investments that pay more typically carry more of it, so a pitch that pairs high returns with safety contradicts that basic trade-off. The SEC's investor site lists guaranteed returns and risk-free offers among its red flags of investment fraud. Many fraudsters work hard to convince investors that very high returns are guaranteed, and the SEC's answer is that such guarantees do not exist.

A Ponzi scheme pays existing investors with money from new ones; in many such schemes the organizer does not invest the money at all, so it needs a constant inflow and tends to collapse when recruiting slows or many people cash out. The SEC flags returns that stay positive whatever the market does, secretive strategies, and difficulty withdrawing. Promoters sometimes offer even higher returns to keep investors from cashing out.

A pyramid scheme pays earlier members with money from new recruits, so income depends on recruiting rather than retail sales, and because each layer must recruit several more, it eventually collapses. It can look like legitimate multi-level marketing, but the FTC says a legitimate one pays you for sales to retail customers without requiring you to recruit. Required inventory purchases and repeated fees are further warning signs.

Promoters spread false or misleading claims, often online and often hinting at inside information, to push a stock up, then sell their own holdings; when the hype stops, the price typically falls. The SEC says scarce public information on small microcap stocks makes false claims easier to spread, and their lower liquidity makes their prices easier to manipulate. A promoter's disclosure of being paid does not make the promotion legitimate.

The SEC says much of the investment fraud in the U.S. is committed by people who are neither licensed nor registered, and most Ponzi schemes involve unlicensed sellers or unregistered firms. Registration gives investors access to information about a company's management and finances. Fraudsters may offer their own proof, but the SEC does not issue registration certificates, and a Form D filing does not show registration.

The SEC lists secretive, complex strategies among Ponzi warning signs and urges caution when a program's commission structure cannot be understood. It notes that some pyramid promoters choose fancy-sounding products that are hard to value, which makes a bogus scheme harder to prove. Its rule of thumb is never to invest in something you do not understand, so lingering confusion is a reason to think twice.

The SEC advises doing independent research rather than just asking the seller for more information or references, since a fraudster gains nothing by correcting you, and checking the salesperson even if you know them socially. Brokers' and advisers' disciplinary histories can be checked for free online. Pressure to invest right now is on its list of red flags, and it says there is never a reason to make a quick investment decision.

The SEC describes advance-fee fraud as a fee, tax or commission demanded up front to release proceeds, and whatever the label, the premise is that you must give money to get money. The FTC applies the same test to loans, credit cards, debt relief and jobs, saying no legitimate business charges you for a mere promise. In relationship investment scams, victims are often persuaded to pay more to unlock an account than they ever deposited.

Advance-fee loan scammers advertise loans or credit no matter your credit history, then demand a processing, insurance or similar fee first, and there is no loan. The FTC says they target people with bad credit, and many buy lists of people who searched or applied online for payday loans. Real lenders check credit before a firm offer, and while some charge application fees, none says a fee guarantees a loan.

Prize scams announce a win and then ask for taxes, shipping, processing fees or account details before the prize can be released. The FTC says real sweepstakes are free and decided by chance, that charging you to improve your odds is illegal, and that scammers pose as government agencies or well-known companies. The notice may feel personal, but the same message usually goes to many people.

Job scammers post ads alongside real ones, may imitate real companies, and can even hold remote interviews before an offer. The FBI says they may then ask new hires to pay up front for background checks, training or start-up equipment, often promising reimbursement in the first paycheck. The FTC says no honest employer charges people to get a job or sends a check to deposit and asks the new hire to send part of it on.

The scammer sends a check, usually for more than is owed, with a story for why you must send some of it back or on to someone else. Banks must make deposited funds available quickly, but a fake can take weeks to be discovered, and by then the scammer has the money you sent and you are left repaying your bank. The FTC advises not relying on a check's funds unless you know and trust the person.

Recovery scammers buy lists of people already scammed and offer, for a fee, to get lost money back, sometimes posing as a government agency; the FTC says agencies never charge to help with a refund. It also says debt settlement companies cannot collect fees before settling a debt and that charging upfront for mortgage relief is illegal. Applying for federal student loan relief programs is free.

The FTC says real prizes are free and no honest employer charges people to get a job. Lenders are a partial exception, since real ones may charge an application or appraisal fee, but no legitimate lender says a fee guarantees a loan. When unsure, the FTC advises finding a company's contact information yourself rather than trusting what the person who contacted you provides.

A 2022 systematic review found no evidence that older adults make up a larger share of online fraud victims than other ages, no consensus on how cognitive function affects who is deceived, and unclear links with trust. The National Institute on Aging says abuse can happen to any older adult, though people without family or friends nearby, with memory problems, or who depend on others for daily care may be more vulnerable.

The CFPB notes that exploiters can be strangers who gain an older person's trust, but also family members or friends. In 1,238 elder abuse cases reported to a UK charity helpline, most involving money, family members were the alleged abusers in about 79 percent, most often adult children. The cases are helpline reports, mostly made by relatives, so the figures describe calls, not population rates.

A financial power of attorney names an agent to handle money matters, and the CFPB notes it grants broad authority without regular oversight. Misuse includes spending the money on the agent, making unauthorized gifts or beneficiary changes, and pressuring for more authority. Safeguards the CFPB describes include telling others about the arrangement, requiring the agent to report transactions to another person, and revoking it.

The CFPB lists a relative, caregiver or friend who blocks visitors or phone calls, will not let the older person speak for themselves, or seems to control their decisions among signs of financial exploitation. The Justice Department describes a new friend who becomes steadily more involved until they control much of the person's life. The NIA counts keeping someone from close friends and relatives as emotional abuse in itself.

In a Chicago study of 420 older adults without dementia, faster cognitive decline over an average of 5.5 years predicted poorer decision making and higher self-reported scam susceptibility, even in those with no cognitive impairment. A Japanese study of 101 older adults tied mild decline to higher vulnerability but found no significant link with executive function. Both are correlational, and neither shows a loss of capacity to consent.

The CFPB's list of signs includes missing money or property, unexplained withdrawals or large wires, usually paid bills going unpaid, names added to accounts that the owner cannot explain, changed beneficiaries, and unusual gifts to a new best friend. The NIA advises watching for a pattern and says someone who reports a concern does not need to prove abuse, since professionals will investigate.

Brokerage firms ask customers to name a trusted contact person they may call if they suspect financial exploitation or cannot reach the customer; that person gets no authority over the account. A power of attorney can require the agent to report transactions to someone else. The SEC's investor site also advises reviewing account statements regularly for any activity you did not authorize.

Timeshare promoters often say an offer will be gone tomorrow, and the FTC suggests asking why, since the company sets its own prices. It calls the rush a warning sign that the seller may want to keep buyers from consulting trusted advisers or taking time to think. The CFPB likewise advises watching out for deals that are only good today.

The FTC describes timeshare presentations where buyers wait a long time between stages or meet with several people in turn, a tactic it says is meant to exhaust them until signing seems the quickest way out. The FTC suggests taking all the documents away and reviewing them alone or with someone trusted before committing.

A lower monthly payment can come from a longer loan, which means paying more interest over its life, so the CFPB says the total cost matters more than the payment many buyers focus on. The FTC adds that low monthly offers often require longer terms and higher rates. A longer car loan also raises the risk of owing more than the car is worth for a longer stretch.

In a 2024 case, the FTC and Arizona alleged that a car dealer slipped in add-ons such as theft protection and paint coating that buyers had not agreed to, and often said they were required when buyers asked to remove them. The CFPB notes add-ons are optional and negotiable, and financing them raises both the monthly payment and the amount borrowed. The FTC advises asking a dealer to slow down when terms move quickly on a tablet.

The FTC warns that free trials often turn into paid charges if not canceled on time, that pre-checked boxes can authorize ongoing charges, and that unclear cancellation terms may mean canceling will be hard. Its home security guidance adds questions about contract length and early-termination penalties, noting that a free alarm can require a long and expensive monitoring contract.

The rule covers some sales at your home, workplace or dorm, or at temporary sites such as hotel rooms and fairgrounds, and the seller must tell you about the right to cancel and provide cancellation forms. It does not cover sales made entirely online, by mail or by phone, or sales of real estate, insurance or securities, among other exclusions. Some state laws give more rights.

The CFPB advises asking for information in writing and walking away from any deal that leaves no time to read the details, ask questions or get advice, and it calls being told to pay right away a warning sign. The FTC suggests taking presentation documents home to review alone or with someone you trust, and the CFPB notes that car buyers can ask for loan disclosures early and take them home to compare offers.

Researchers note gaslighting is defined inconsistently and often blurred with broader manipulation. A 2025 systematic review of nearly 100 sources on gaslighting in gender-based violence proposed defining it as manipulation of perception, reality and memory that undermines a person's ability to know their world, shown in tactics and effects. Most of that work concerned intimate partners, and most came from English-speaking Western countries.

In a 2025 review of the gaslighting literature, denying or dismissing abuse appeared in about 60 percent of the publications reviewed and challenging the target's memory in about 42 percent. Publications quote survivors whose partners insisted they had said things they knew they had not, or accused them of acts they knew they had not done. Disrupted memory and doubt about one's own experiences were among the effects reported.

A 2025 review found that misrepresenting situations to create false narratives appeared in about 62 percent of the gaslighting publications it examined. Victim blaming and role reversal, in about 46 percent, recast the target as the one at fault, as when a partner accused the target of the cheating the partner was doing. One quoted account describes a woman who began to doubt whether she had been pushed, because her partner seemed so certain.

A 2025 review found that making the target feel crazy, or calling them crazy, was described in about 58 percent of gaslighting publications, and minimizing or justifying abuse in about 42 percent. Labels such as crazy, emotional or irrational attack the target's standing as someone who can know what happened. One workplace framework treats trivializing an employee's concerns as a necessary feature of workplace gaslighting.

A 2025 review found isolation from support systems described as a gaslighting tactic in about a quarter of publications, with the effect of making targets more dependent on the abuser. Publications also describe convincing targets that no one would believe them, and one suggests that social isolation may have fed a sense of losing touch with reality. Most of this work is qualitative and descriptive, not a test of cause.

Building on earlier work, a 2025 review held that questioning someone's view counts as gaslighting only when it produces effects such as self-doubt, which separates it from ordinary disagreement or dismissal, and warned that the term risks becoming overgeneralized. Scholars disagree about intent: some say gaslighting can happen without awareness, and a workplace framework treats intent as possible but not required.

A few publications in a 2025 review describe survivors recording conversations; one quoted survivor began after her partner kept saying she had said things she knew she had not. In an Australian survey, 37 women who had experienced partner sexual violence wanted service providers to believe them and help name it as violence, a theme the researchers called counteracting the gaslighting. Neither finding tests whether a strategy works.

Researchers use DARVO for a response in which the accused denies wrongdoing or harm, attacks the accuser's credibility, and casts themselves as the real victim of a false accusation. A 2024 paper summarizes an earlier survey in which nearly 72 percent of 138 students who had confronted someone reported all three moves. The idea began with sex offenders, but those students' reports ranged from betrayed secrets to abuse.

In an experiment summarized in a 2024 paper, 230 students read a fictional perpetrator's account of an assault, with or without DARVO; those who saw DARVO rated the perpetrator more believable and less responsible, and the victim less believable. In five later experiments, labeling an accused person the victim shifted support, but reliable effects appeared only among readers who said that wording influenced them.

Two studies summarized in a 2024 paper, one of students who had confronted someone and one of college women after sexual assault, found that the more DARVO people experienced, the more they blamed themselves. Separate research on sexual violence survivors links self-blame to not disclosing and to higher PTSD risk. These are associations from self-reports and cannot show that DARVO alone causes self-blame.

Institutional betrayal is an organization harming people who depend on it, such as by covering up a report or punishing the reporter. A review of campus sexual violence research lists disbelief, victim blaming, discouraging reports and punishing whistleblowers among contributing responses. In an online survey of 317 workers who faced workplace sexual harassment, about 55 percent reported betrayal, which was linked to lower job satisfaction.

Taxonomies of propaganda and deception group whataboutism with the red herring and the straw man as distraction techniques: it pulls attention from the main issue by shifting blame or criticism onto others. A study of UK newspaper editorials found it used to argue that other nations' emissions reduced the UK's own responsibility. Whatever the other party's faults, the move does not address the original claim.

Tone policing restricts which ways of voicing a complaint count as acceptable. An argumentation scholar lists it among the devices a more powerful party can use to suppress the justified objections of a less powerful one, and an ethics essay argues that demands for calm, professional speech let shouting or crying be read as a sign that the complainer is the problem. Both are conceptual analyses, not experiments.

A 2026 paper notes earlier findings that awareness of DARVO reduced its impact, and a 2024 study suggested that advocates learn the pattern so they can identify and interrupt it. Recognizing it means noticing when a dispute about a specific act has turned into a dispute about the credibility of the person who raised it. The 2024 suggestion is a recommendation, not a tested result.

Evan Stark describes coercive control as a course of conduct rather than separate events, with physical violence often, not always, among its tactics. Its harm lies in combination and repetition: indirect tactics such as isolation and microregulation erode independence while intimidation compels compliance. Treating abuse as separate assaults misses the control in between, and single acts may not look like violence.

Federal guidance lists forcing isolation from family, friends, school or work among forms of psychological abuse. In one study of 15 Australian women, isolation came through guilt over social contact, forced moves and disapproval; one stopped making plans because defending each friendship was too tiring. A rural Canadian study adds that distance favored abusers and poor phone service made help hard to reach.

The Justice Department defines technological abuse to include using devices, apps or location trackers to monitor or control a partner. Women in one small Australian study described GPS tracking and hacking that left them feeling constantly watched. A review of qualitative studies of young people found that password sharing offered as proof of trust could also serve as a way to spy and control.

Federal guidance defines economic abuse as restricting someone's ability to get, use or keep money, assets, credit or financial information, including forcing default on debts. Rural survivors in one Canadian study described dependence built gradually in ways outsiders would not notice, from interference at work to a monthly allowance that left one woman unable to leave the farm.

Researchers list microregulation among the indirect tactics of coercive control that erode independence and build dependence on an abusive partner. In survivor accounts it looked mundane: one partner capped how many hours a woman could work, another took her car just before she had to leave, and control of money narrowed daily choices. Each act can seem minor; researchers point to the cumulative distress of tactics that are not openly harmful.

Survivors in one small qualitative study described humiliation over housework, appearance and sex alongside accusations and micro-management, and spoke of dread and walking on eggshells. Researchers describe coercive control as creating continuous anticipation of negative or unpredictable consequences, with swings between idealization and denigration deepening dependence.

Recognizing coercive control does not make leaving automatically safe, since abuse can continue or intensify after separation. A review of femicide risk tools notes research agreement that escalating violence, separation and death threats are among the main risk factors. Such tools are built to inform safety planning, and federal resources point to a national hotline with advocates.

Lenore Walker described three phases: tension builds, an acute incident occurs, and a honeymoon phase follows in which the abuser seeks forgiveness. The model describes a repeating pattern, but it has limits: CDC notes partner violence ranges from one episode to chronic abuse over years, and critics argue the syndrome built on it explains abuse through the victim's psyche rather than the violence.

In Walker's model, the reconciliation phase leads the victim to believe the abuser still cares and deserves another chance, which can keep them from leaving. Research on separation describes a nonlinear process with temporary breakups and reconciliations; money and housing problems can prompt a return, and survivors often make several attempts to leave.

Federal guidance counts threats to harm a partner's children, family or friends, and destroying pets or property, as psychological abuse. Rural survivors in one Canadian study reported threats against their children, livestock and companion animals, some involving firearms. Research reviewed in that study found that people with companion animals sometimes delay leaving because they have nowhere to go with them.

Federal guidance defines technological abuse broadly: any use of devices, platforms, cameras, apps or trackers to harm, control, stalk, impersonate, exploit, extort or monitor someone. A review of studies of young people found password sharing framed as trust, pressure to sext and sharing sexual photos without consent, and noted that blocking an abuser could provoke further harassment.

Reproductive coercion and abuse covers pressure toward pregnancy, such as contraceptive sabotage, and pressure against it, such as coercion to end a pregnancy. In records from more than 5,000 Australian pregnancy-counseling clients, counselors noted it commonly, and about equally often in each direction. Federal guidance also counts denying a partner medical care as physical abuse.

Research reviewed in a 2022 study links delayed separation and returns to limited money, housing problems and fear for children's safety, and notes that abuse can continue or intensify after leaving. Migrant survivors may also face insecure legal status and financial dependence. Critics argue that approaches such as battered woman syndrome cast survivors' responses as irrational, erasing the violence and their own judgments of danger.

The Justice Department's violence-against-women office does not serve the public directly; it points people to a national hotline, to advocates by chat at any hour, to local programs, and to 911 in an emergency. Evidence is growing that survivors weigh risks on their own, but there is no consensus on how accurate those judgments are; structured risk tools are designed to support safety planning.

Researchers describe joining an abusive group as a sequence of decisions, such as an introductory meeting, then a more intensive seminar or residential stay, each narrowing the distance to the group's authority. The first offer is often warmth, idealism and shared mission. Evidence does not support the idea that all members had a preexisting disorder; life transitions and unmet needs for belonging play a part.

In a 2026 survey of Italian psychologists, about a quarter of respondents named love bombing, meaning fake love or an intensely welcoming attitude used to attract and retain members, as a feature of abusive groups. A forensic interview tool for group manipulation lists it among recruitment maneuvers, alongside rewarding loyalty and punishing or excluding members who break the rules.

One manipulation taxonomy defines information control as managing the information and communication devices members use so the group monopolizes what reaches them. Psychologists surveyed in Italy rated attacking critics with ad hominem arguments as characteristic of abusive groups. Control need not be total: in one case analysis members had smartphones, but a leader steered which sources they used.

A forensic interview tool for group manipulation includes control over language, meaning an altered vocabulary and jargon unique to the group, and denigration of critical thinking, which supplies clear, final answers to members' doubts. One propaganda-annotation scheme files the thought-terminating cliche under simplification, as a device for shutting down critical thinking. These are classification schemes, not experiments.

Researchers describe group commitment growing through progressively larger demands, so the immediate cost of resisting feels higher than one more small step, until even large donations feel freely chosen. Leaving can bring grief for lost relationships and identity, isolation, economic instability and stigma. Former members also report relief and freedom, so exit is mixed rather than only a loss.

A 2015 list of misleading psychology terms flags brainwashing: so-called brainwashers use standard methods such as commitment, manufactured credibility, staged consensus and vivid testimonials. Of about 3,500 American prisoners indoctrinated during the Korean War, only a small minority falsely confessed and probably under 1 percent kept the ideology. A 2026 group-abuse study calls the label outdated.

The strategies described in group-abuse research include isolation, control of information and control over members' finances and personal lives. On the other side, researchers list maintained family ties, support from significant others, outside social networks and access to therapy as protective factors in leaving and recovery. These come from reviews and surveys, not tested interventions.

Workplace bullying research separates work-task attacks from social and personal ones; task attacks include withholding information, repeated reminders of faults, intensive inspection and intentional changes to someone's tasks that make the work hard to complete. A standard questionnaire also asks about tasks with unreasonable or impossible targets or deadlines. One such act is not bullying; the definition requires persistence.

A 2021 study's overview of bullying gives taking credit for others' work and making unreasonable demands as examples of work-related bullying. The Negative Acts Questionnaire asks how often, over six months, someone had allegations made against them or was repeatedly reminded of errors. Its items avoid the word bullying and count specific acts, so it measures a pattern over time, not an overall impression.

The Negative Acts Questionnaire, used to measure workplace bullying, asks how often someone withheld information that affected the respondent's performance. Bullying research describes this as a work-task attack: it makes the work difficult to complete and so harms the target. Researchers count it as bullying only when it persists, and the questionnaire asks about the previous six months.

Bullying questionnaires ask how often someone was given tasks with unreasonable or impossible targets or deadlines, or exposed to an unmanageable workload. Bullying research also describes workers forced to accept unreasonable deadlines or improperly forced to change their tasks. A single crunch is not the pattern; the research definition requires persistence over time against someone who finds it hard to push back.

A common research definition describes bullying as persistently receiving negative actions over a period of time from people the target finds it hard to defend against, and explicitly excludes one-off incidents. Questionnaires such as the Negative Acts Questionnaire avoid the word bullying and ask how often specific acts occurred in the past six months, so the pattern is what gets measured.

In one set of studies, making a job seem more meaningful lowered the minimum salary participants would accept, and employees with more meaningful work turned down better-paid offers more often. That trade can be a free choice, and the studies did not test manipulation. Separately, bullying questionnaires list being pressured not to claim sick leave, holidays or expenses one is owed.

In a study of junior faculty in India, bullying was linked to speaking up when staff saw their organization as intolerant of bullying, and to disengagement when they saw it as tolerant. An intervention that redesigned people-management practices was associated with reduced bullying risk, partly through changes in job demands, shifting prevention from individuals to the system.

FTC guidance lists pressure to act immediately as one of four signs of a scam, notes that callers may tell people not to hang up so they cannot check the story, and observes that honest businesses allow time to decide. In lab studies, people judging emails or fraud materials under time pressure made more errors, though one study found this only for scams framed as avoiding a loss.

One addiction-treatment protocol teaches a firm no, at most a short reason without self-justification, then the broken record step of adding no further explanations. An Irish training program for doctors refusing benzodiazepine requests has them repeat the same apology, reason and alternative offer as often as needed. Neither program tested the technique on its own.

One assertiveness framework calls saying no a foundational skill, and one refusal protocol teaches a firm no with a short reason given without self-justification. Consumer guidance adds that honest businesses give people time to decide and that anyone pressing for payment or personal information is a scammer. The same framework notes that the most effective style varies with culture and context.

Implementation intentions are if-then plans formed ahead of time. A meta-analysis of 16 studies found they reduced weekly drinking slightly, more in community samples, with no clear effect in university samples or on heavy drinking episodes. One refusal program rehearses lines in role-play, but none of this work tested scam calls.

Asch's effect depended on unanimity: later accounts of his studies report that, against a majority of six, a single dissenter cut conformity from about a third of answers to about 5 percent. A 2021 study found that a dissenting minority's expertise mattered only when siding with the majority cost nothing. Consumer guidance makes a related point: telling a friend, relative or neighbor before acting can help someone see a scam.

One assertiveness framework describes I-statements, such as I think, I feel or I want, as building a habit of stating honest thoughts and feelings directly, and credits a 1970s training book with framing saying no as a way to set boundaries. Its functional assertiveness idea stresses persistence and a style fitted to context and culture, not one forceful script. It is a conceptual framework, not a trial.

In Patterson's coercive family process model, a child escalates to escape a demand and a caregiver withdraws the demand to avoid conflict, so each reinforces the other's behavior. One parent-training program teaches consistent, mild sanctions such as time-out, yet a Dutch trial found it no better than usual care, with both helping. The evidence is from families; with an abusive adult partner, a consequence such as leaving can raise risk.

FTC guidance sorts next steps by payment method: the bank for transfers and debit cards, the card issuer for credit cards, and the app, wire company, gift card issuer or crypto exchange for those payments. Federal law limits losses from unauthorized card use, while cryptocurrency lacks the same protections. Mailed cash can sometimes be intercepted before delivery, for a fee.

The FTC takes fraud reports and uses them to build cases, spot trends and educate the public. The FBI's Internet Crime Complaint Center collects cyber-enabled crime reports, shares them with field offices and partners, and in some cases helps freeze stolen funds, though it cannot respond to every complaint. A federal online tool matches a scam to the right agency, and state attorneys general also take reports.

The FBI's complaint form asks for account details, transaction dates and amounts, who received the money and what is known about the person responsible. Investigators may later request receipts, phone bills, emails with full headers, web pages and chat logs, and the complaint center does not take attachments, so originals stay with the person reporting. FTC guidance also says to keep gift cards and receipts.

A credit freeze stops anyone, including the owner, from opening new credit until it is lifted; it is free and does not affect credit scores. A fraud alert instead asks lenders to verify identity before granting credit. For a compromised account, FTC guidance is to set a new strong password, change it anywhere it was reused, and turn on two-factor authentication.

The FTC describes scammers buying and trading lists of people who have already paid a scammer, then posing as a government agency, law firm or advocacy group that can recover the money for a fee. Government agencies do not charge to help get refunds. The FBI's complaint center says it never contacts people directly for money and does not work with private firms to recover funds.

A 2026 systematic review of 21 studies found fraud victimization associated with shame and self-blame, and found that harm tracked manipulation and betrayal more than the size of the loss. Fear of judgment and stigma were often reported as barriers to disclosure and seeking help, which limits access to support. The authors call for more longitudinal and intervention research.

Susceptibility shifts with circumstances: in lab studies, time pressure alone made people worse at detecting fraud. A 2026 review notes evidence that dismissive or blaming responses from institutions such as banks and law enforcement may deepen distress and discourage help-seeking. Reports matter beyond the individual, since the FTC and FBI use them to build cases, track trends and warn others.

A 2012 replication of the elderly-priming study found no slower walking with automated timing; the effect appeared only when experimenters expected it. Preregistered replications of ego depletion with 2,141 and 3,531 participants found effects near zero. Results run both ways: a 2023 Asch replication found an error rate of about a third, close to the original.

In most Asch-style experiments a wrong, conforming answer costs nothing, which raises questions about external validity. When a 2023 replication paid for correct answers, the error rate fell from 33 to 25 percent, and earlier incentive studies were mixed. Fraud studies face a similar gap; some measure susceptibility with rating tasks that stand in for real victimization.

James Vicary claimed in 1957 that flashed messages raised popcorn and cola sales, then admitted fabricating the data. Meta-analyses put subliminal persuasion effects around a correlation of 0.06, with lab effects appearing mainly when the primed brand fits a current need, such as thirst. Whether such effects carry over to real advertising remains strongly debated.

A large nudging meta-analysis reported an average effect of d = 0.43, reduced to 0.31 after a moderate bias correction; critics argued fuller correction leaves little or no effect. In 126 nudge-unit trials free of selective publication, the average effect was 1.4 percent versus 8.7 percent in published studies. Effects vary widely by type, so some nudges may still be large.

In a 2019 phishing study, cognitive reflection and sensation seeking predicted susceptibility only modestly, while time pressure made errors more likely. A review of older adults found no evidence they are overrepresented among online fraud victims and no consensus on the role of cognitive decline. A 2025 study linked self-reported financial exploitation to combinations of socioeconomic status, emotion regulation and health.

In a field experiment with over 10,000 Dutch ministry staff, 22 percent of the control group gave away a password; an information campaign cut that by about 6 points and one simulated phishing email by about 8, with no clear gain from combining them. Hospital data link repeated simulations to fewer clicks. The authors note technical safeguards help but are rarely enough alone.

A review of 1,149 persuasion studies found most message choices change persuasiveness only slightly, with prediction intervals spanning both helping and hurting. Published nudge studies reported effects about six times larger than unselected agency trials, and a priming replication showed experimenter expectations could produce an effect. Asking who, against what control, measured how, and replicated where exposes these gaps.