The 2026 evidence does not support a choice between trusting every opportunity and distrusting entrepreneurship. Independent work is widespread, household pressure is real, and AI is already a legitimate workplace tool. At the same time, reported losses are substantial, social platforms are an important entry point for harmful offers, and recent enforcement records repeatedly involve unsupported earnings, unclear affiliations, refund representations, and incomplete cost pictures. The practical response is not to abandon the goal. It is to require decision-grade evidence before money, time, debt, or dependence increases.
What this report measures - and what it does not
This is a baseline evidence review. It combines publicly available government data, a nationally representative household survey, current regulatory material, and selected public enforcement records. It does not use private PauseThePitch customer reports, and it is not a survey of every course, side hustle, coaching program, or online business offer.
The numbers in this report come from different evidence systems. Federal Trade Commission figures are based largely on reports submitted by consumers and data contributors; they are not a prevalence survey and likely do not capture every loss. Federal Reserve figures come from the Survey of Household Economics and Decisionmaking, a probability-based survey of nearly 13,000 U.S. adults. Census Bureau nonemployer figures come from administrative and census data. Enforcement complaints contain allegations unless and until resolved; settlement announcements describe the terms and allegations identified by the agencies involved.
Interpret the numbers at their actual scale
The FTC's approximately $16 billion in reported 2025 fraud losses covers many kinds of fraud, not merely business opportunities. The $2.1 billion social-media figure also spans multiple scam categories. These figures establish the scale and distribution channel of reported harm; they do not establish that ordinary social advertising, coaching, independent work, or AI tools are inherently deceptive.45
The 2026 context: legitimate ambition under real pressure
Opportunity pitches do not appear in a vacuum. In the Federal Reserve's October 2025 household survey, just over nine in ten adults said price increases were at least a minor concern, and 42% said finding or keeping a job was at least a minor concern, up from 37% the year before. The same report found that one in four workers had used generative AI at work in the prior month.1
Independent earning is also a real part of the economy. The Census Bureau reported 29.8 million nonemployer businesses for 2022, the latest demographic release used here, with $1.7 trillion in receipts. These are businesses without paid employees that meet the program's tax-filing and receipts criteria.3 The Federal Reserve separately found that 20% of adults performed some kind of gig activity in the prior month during 2024. Most gig activity was part-time, and gig workers reported less financial security than adults who did no gig activity.2
Those facts matter because skepticism can be framed badly. A person considering a service business, course, online store, freelance skill, or second income is not necessarily chasing fantasy. The goal may be practical. The question is whether the particular route being sold has evidence proportionate to its cost and risk.
Seven findings from the current evidence
Financial pressure is part of the market for opportunity pitches
When prices, job security, debt, retirement, or scheduling are active concerns, a clear income story becomes more relevant. Relevance is not proof. A pitch can accurately recognize a household problem and still provide weak evidence for its solution. The first useful separation is between the buyer's durable goal and the seller's replaceable offer.
Buyer implication: Write the goal without the product name. If the goal survives but the offer changes, comparison becomes possible.
Entrepreneurship is too real to treat every offer as suspect
Tens of millions of nonemployer businesses and widespread gig activity show that earning outside a traditional job is neither rare nor automatically implausible. However, business counts and receipts do not prove the viability of a particular method. Receipts are not profit, and a large market does not guarantee that a beginner can enter economically.
Buyer implication: Ask for unit economics, customer acquisition, recurring work, capacity limits, and net income after expenses rather than relying on the broad popularity of entrepreneurship.
Social discovery deserves its own verification step
The FTC reported that nearly 30% of people who reported losing money to a scam in 2025 said it began on social media, with $2.1 billion in reported losses across scam categories. The agency said social platforms allow targeting, impersonation, advertising, and rapid movement into private messages or groups.5
This does not mean a social-media offer is false. It means the platform that introduced the offer should not also be the only place used to verify it.
Buyer implication: Leave the feed. Search the legal business name, seller history, independent complaints, applicable licenses, court or regulator records, and the exact earnings language outside the seller's community.
Reported harm in business and job opportunities is material
The FTC's latest detailed annual category release reported $750.6 million in 2024 losses associated with business and job opportunities, almost $250 million more than in 2023. Within that category, reported losses involving job and employment agency scams rose from $90 million in 2020 to $501 million in 2024.6
Reported-loss data should not be used as a failure rate for legitimate opportunities. It does establish that the category warrants more than an intuitive credibility check.
Buyer implication: Preserve the original advertisement, sales page, webinar, contract, refund terms, receipts, financing, and written promises before they change or disappear.
AI is both a real tool and a borrowed credibility signal
Generative AI is genuinely used at work: one in four workers in the Federal Reserve survey reported using it during the prior month, and most users said it saved time.1 That makes AI a plausible operational tool. It does not make an AI-labeled income system profitable by default.
Recent FTC matters involving ecommerce and business-growth offers have included allegations or findings around AI-powered systems, passive-income representations, earnings claims, affiliations, and refunds. The issue is not the presence of AI. It is the unsupported jump from a useful technology to a likely financial outcome.
Buyer implication: Ask which exact task the AI performs, what human work remains, what the output replaces, what it costs at production volume, and what evidence ties that function to net customer results.
The same proof gaps recur across different offer types
Recent public matters do not represent the entire market, but they reveal a practical pattern. The questions repeatedly concern how earnings were supported, whether well-known affiliations were real, whether refund or buy-back representations matched practice, whether required disclosures were provided, whether later costs were visible, and whether truthful negative reviews were constrained.
The FTC's pending rulemaking is also evidence of the policy question, not a final rule. In January 2025, the Commission proposed expanding parts of the Business Opportunity Rule to additional money-making opportunities, including some coaching and investment offers, with attention to material misrepresentations, substantiation, and records.7
Buyer implication: Treat every earnings example as a claim requiring a denominator, time period, expense definition, participant characteristics, and written basis.
The safest commitment is the next one the evidence has earned
A decision does not need to be all-in or all-out. A reasonable next step may be a customer interview, a manual pilot, a small inventory test, one month of software, an independent skills course, or no purchase until documents arrive. Escalating because of prior effort, community pressure, a deadline, or money already spent is different from escalating because the next stage has evidence.
Buyer implication: Set cash, time, debt, privacy, and reputation caps before purchase. Define what must happen before another payment and what result triggers stopping.
What current enforcement records signal
The cases below are included because they expose recurring evidence questions. They are not a random sample and cannot establish how common a practice is across the whole market.
| Public matter | Current status used here | Buyer-facing evidence question |
|---|---|---|
| Air AI | March 2026 FTC settlement announcement and proposed order involving allegations about earnings, business growth, refund or buy-back guarantees, disclosures, and refunds.8 | What written evidence supports the earnings and refund representations, and do the actual conditions match the sales impression? |
| IM Mastery Academy | May 2026 FTC and Nevada settlement announcement resolving charges involving allegedly false or baseless earnings claims used to sell financial training and an MLM venture.9 | What generally happened to customers or participants after costs, and how were featured earnings selected? |
| Click Profit | August 2025 FTC settlement announcement following allegations involving guaranteed passive income, AI claims, purported brand affiliations, and restrictions on truthful reviews.10 | Can platform, supplier, technology, and brand relationships be verified independently, and can customers speak freely? |
The 2026 decision-grade evidence request
A buyer does not need to prove fraud to decide that an offer has not earned a larger commitment. The following request works across many courses, coaching programs, packaged businesses, automation offers, and income systems.
- Claim: State the exact financial, time, customer, or performance claim without motivational language.
- Population: Identify how many relevant purchasers started, completed, attempted implementation, and produced the displayed result.
- Time: State the measurement period and the time participants spent before and during the result.
- Costs: Include the purchase, financing, software, traffic, inventory, contractors, fees, taxes, refunds, and operating reserves needed through a valid result.
- Outcome: Separate revenue, gross profit, owner compensation, cash flow, and net profit. State whether results are typical, selected, or exceptional.
- Dependencies: Identify the platform, supplier, account, license, advertising channel, territory, algorithm, coach, or third party the model depends on.
- Terms: Provide the contract, refund and cancellation process, data access, financing consequences, review restrictions, and dispute terms before payment.
- Fit and stopping rule: Explain who should not buy, what commonly prevents success, and what evidence should cause a buyer to stop.
The standard is proportional
A $25 book does not need the same diligence as a financed $25,000 business package. The amount of evidence should rise with price, debt, irreversibility, operating exposure, and dependence on the seller.
What the next update should watch
The next edition should update the evidence rather than merely change the date. Four developments deserve attention:
- New 2026 complaint and loss data. Use the most recent FTC category detail when released, while preserving the distinction between reports and prevalence.
- Rulemaking status. Track whether proposed money-making-opportunity and earnings-claim requirements advance, change, or remain pending.
- AI claim language. Distinguish measurable automation from broad assertions that AI makes revenue passive, guaranteed, or unusually accessible.
- PauseThePitch's future original dataset. Develop a privacy-reviewed, anonymized classification method before publishing any platform-derived trend. Until that process exists, do not imply proprietary findings.
A future quarterly comparison should publish the same definitions, show what was added or revised, and avoid claiming a trend from a change in source coverage alone.
Methodology and limitations
- Research question
- What does current public evidence support about the environment in which online money-making pitches are evaluated, the channels associated with reported harm, and the proof gaps a buyer can test?
- Evidence window
- Sources published from January 2025 through August 8, 2026, using underlying data that primarily cover 2022 through 2025.
- Included evidence
- Federal Trade Commission reports, consumer guidance, rulemaking, and enforcement announcements; Federal Reserve household research; and Census Bureau nonemployer-business data.
- Excluded evidence
- Unverifiable social posts, vendor-sponsored market-size claims without disclosed methods, anonymous anecdotes, private customer content, and claims that could not be traced to a primary source.
- Analytic method
- Structured evidence review. Each quantitative statement was tied to its measured population and period. Enforcement allegations were not treated as market prevalence or automatically as adjudicated facts.
- Primary limitation
- This is not an original prevalence study and does not estimate the share of legitimate, weak, deceptive, or fraudulent offers. The selected enforcement matters illustrate proof questions, not the frequency of conduct.
PauseThePitch is an educational decision-support service, not a regulator, law firm, investment adviser, or credit counselor. This report does not make legal findings about any unreviewed seller and should not replace qualified professional advice where the stakes require it.
Sources
- Board of Governors of the Federal Reserve System, Economic Well-Being of U.S. Households in 2025, published May 2026. Nationally representative SHED fielded in October 2025.
- Federal Reserve, Employment and Gig Work, published May 2025 using October 2024 survey data.
- U.S. Census Bureau, Demographic Characteristics of Nonemployer Business Owners, published May 2025 using 2022 data.
- Federal Trade Commission, 2025 Reported Fraud and Imposter Losses, published June 2026.
- Federal Trade Commission, Reported Losses to Scams Starting on Social Media, published April 2026.
- Federal Trade Commission, 2024 Consumer Sentinel Loss Highlights, published March 2025.
- Federal Trade Commission, Proposed Business Opportunity Rule Amendments, January 2025. Cited as a proposal, not a final rule.
- Federal Trade Commission, Air AI Settlement Announcement, March 2026.
- Federal Trade Commission, IM Mastery Academy Settlement Announcement, May 2026.
- Federal Trade Commission, Click Profit Settlement Announcement, August 2025.
- Federal Trade Commission Consumer Advice, Business Offers and Coaching Programs, accessed August 8, 2026.
Citation note: Dates describe publication or access. Underlying measurement periods are stated in the report and should not be read as real-time 2026 market estimates.