A few years ago, defrauding a family of means took real work. The criminal had to research the target for weeks, learn the family’s relationships and habits, and then execute an impersonation convincing enough to move money. That effort protected most households. Con artists chose their marks carefully because every attempt was expensive, and most families were never worth the investment.
AI has removed much of that effort. A voice can now be cloned from a short audio clip lifted from a podcast interview, a conference panel, or a grandchild’s graduation video. A fraudulent email written in your attorney’s exact style takes seconds to produce. A live video call with a synthetic version of a familiar face is no longer a research project. The economics that once protected wealthy families have reversed, and the same visibility that comes with a successful business, a charitable board seat, or a public liquidity event now doubles as targeting data.
That reversal is why we treat household fraud protection as a wealth management discipline in its own right, alongside tax strategy, estate design, and portfolio construction. This article walks through what has changed, the four schemes we believe deserve the most attention in 2026, and the verification habits that can help a family stay ahead of them.
The Economics of Fraud Have Shifted
Americans reported $20.9 billion in internet-crime losses in 2025, a 26% increase in a single year.1 For the first time in the report’s history, the FBI broke out complaints tagged as AI-related, which represented nearly $893 million in reported losses, and that figure almost certainly understates the problem because victims rarely know whether a machine wrote the email that fooled them.1 Americans age 60 and older reported $7.748 billion in losses, up 59% from the prior year.1
The trajectory is steeper than the totals. Generative AI could push fraud losses in the United States to $40 billion by 2027, up from $12.3 billion in 2023, a compound growth rate of 32% a year.2
The 2026 Fraud Landscape in Four Numbers
What U.S. households reported, and where the trend line points
$20.9B
Reported U.S. internet-crime losses in 2025
Up 26% in one year$7.7B
Reported losses by Americans age 60 and older
Up 59% from 2024$893M
Losses in complaints tagged as AI-related
First year the FBI tracked it$40B
Projected U.S. generative-AI-enabled fraud losses by 2027
Up from $12.3B in 2023Sources: FBI Internet Crime Complaint Center, 2025 Annual Report; Deloitte Center for Financial Services. Reported losses understate actual losses because many incidents are never reported.
Affluent families sit directly in the crosshairs, as they hold the balances that justify a criminal’s time, they move large sums routinely enough that a fraudulent transfer request doesn’t look unusual, and their lives generate exactly the raw material modern impersonation requires. A recorded keynote, a charity gala photo caption, a deal announcement in the trade press, a teenager’s social feed. Each one is a training set.
The Four Schemes to Watch in 2026
Fraud categories blur together, but four patterns account for most of the serious losses we see aimed at families with meaningful wealth. Each pairs an old confidence trick with new synthetic media, which is why familiar instincts keep failing against them.
Arrives as a distressed call from a child or grandchild who needs money moved immediately and asks you to keep it private.
Hang up, call the family member back on the number you already have, and use your family verification phrase.
Arrives as a video meeting or voicemail from someone who looks and sounds like your advisor, banker, or attorney requesting a transfer or credentials.
Verify through a separate channel you initiate. No legitimate professional resents a callback.
Arrives as revised wire instructions inside a real email thread during a closing, an escrow funding, or a capital call.
Confirm every new or changed instruction by phone on a known number before any money moves.
Arrives as an exclusive private investment introduced through weeks of patient, personalized contact, often with a polished platform showing steady gains.
Verify registration, custody, and auditability before committing a dollar. Exclusivity is a sales device, never evidence.
The cloned voice
The family emergency call has existed for decades, but it used to fail whenever the impostor didn’t sound right. That defense is gone. A short publicly available audio clip can produce a clone that sounds like your child, and federal consumer regulators have been warning about exactly this upgrade to the scheme since voice cloning tools became widely available.3
The call usually comes at night. The voice says there’s been an accident or an arrest, a lawyer is standing by, and the money must move now. Urgency and secrecy are the two levers, because both are designed to keep you from doing the one thing that defeats the scheme, which is contacting the real family member directly. Older relatives are frequent targets, and the 2025 IC3 data show why families should treat senior fraud as a household risk: Americans age 60 and older reported $7.748 billion in total internet-crime losses across all categories.1
The synthetic video call
If a voice can be cloned, a face can too, and in real time. The defining case involves a finance employee at the engineering firm Arup, who joined a video call with what appeared to be the company’s chief financial officer and several colleagues. Every participant except the victim was a deepfake. The employee wired roughly $25.6 million across 15 transfers before the fraud was discovered.4
The avenues of attack are wide and frightening. Imagine a grieving widow receiving a video message from someone who appears to be her longtime advisor, asking her to confirm her account credentials, or a family office employee joining a call with a synthetic version of the family principal authorizing an urgent transfer. Seeing is no longer verifying, and the families who internalize that earliest will be the hardest to reach.
The redirected wire
Business email compromise remains a workhorse of serious fraud, producing about $2.8 billion in reported losses in 2024 alone.5 The criminal compromises a mailbox belonging to a title company, an escrow agent, a contractor, or an attorney, watches the transaction develop, then sends revised wire instructions at the moment payment is expected, in the right voice and the right thread.
Consider when your household wires seven figures. For example, a property closing in Sarasota, funding a trust, a capital call from a private fund, or the escrow mechanics of a business sale. These are precisely the moments a criminal waits for, because the transfer is expected, the parties are busy, and a change in banking details can read as routine paperwork. An owner working through the mechanics of a sale should treat payment verification with the same discipline as the rest of the process of getting the business ready before going to market, because a compromised escrow instruction can undo years of careful planning in a single afternoon.
The manufactured opportunity
Investment fraud is the largest single category of scam losses, accounting for nearly half of the total, and complaints involving cryptocurrency alone exceeded $11 billion.6 The modern version rarely opens with a pitch. It opens with a relationship, sometimes months of casual, intelligent conversation, potentially even run in part by AI personas that never get tired and never break character. The pitch arrives only after trust is established, usually framed as a private opportunity unavailable to the general public, complete with a professional platform showing steady returns.
Wealthy investors are conditioned to believe the best opportunities are exclusive, which is exactly the instinct this scheme rents. The counter is structural rather than emotional. Real investments have verifiable registration, third-party custody, and auditable statements. If returns can only be seen on the promoter’s own platform, the platform is the product.
The older schemes haven’t retired
Not all scams are high-tech. Bank impersonation calls remain common, in which a caller posing as your bank’s fraud department, sometimes followed by a second caller posing as a federal agent, convinces the target that their accounts are compromised and must be moved to a “safe” account controlled by the criminal. No legitimate bank or agency will ever ask you to move money to protect it. Tech support pop-ups still open the door to remote access on the computers of older relatives. Text messages about unpaid tolls, missed deliveries, and jury duty exist to harvest the credentials that make the larger schemes possible. AI simply improves the grammar and increases the volume. The defenses are the same ones that defeat the sophisticated versions, which is precisely why a written protocol pays for itself across every threat at once.
Why Complex Households Carry Extra Exposure
Wealth adds surface area. A household with multiple entities, properties, and advisors has more channels through which a fraudulent instruction can arrive and more people authorized to act on one. Assistants, bookkeepers, property managers, and household staff all receive requests that look routine. Aging parents may hold significant assets behind a single email password, and research on cognitive health in retirement points to financial judgment as one of the earliest capacities to change, often years before anyone notices. Meanwhile, the youngest generation posts the raw audio and video, making it trivial to clone the rest of the family.
The common structural weakness is a single point of approval. When one person can authorize a large transfer alone, on one channel, under time pressure, every scheme described above has a clear path. The fix costs almost nothing.
Household Verification Protocols
Technology created this problem, but the best defenses are procedural. We encourage families to adopt a short written protocol, brief everyone who touches money on it, and revisit it annually
Five Habits That Defeat Synthetic Fraud
Treat urgency as the red flag
Every scheme above depends on speed and secrecy. Any request that punishes a pause is presumed fraudulent until verified.
Call back on a number you already have
Never verify a request on the channel it arrived through. Hang up and initiate contact yourself using stored contact details.
Set a family verification phrase
A private phrase, never written online, that any family member can request during an emergency call. A clone can't answer what it was never trained on.
Require two approvals above a threshold
No single person, family member or employee, should be able to move a large sum alone. Pick a dollar threshold and hold to it.
Document and report immediately
If money moves, minutes count. Contact the bank to attempt a recall, file at IC3.gov, and notify your advisor so accounts can be locked down.
A written household protocol works only if everyone who touches money knows it, including assistants, bookkeepers, and household staff. Review it annually. For educational purposes.
Beyond the protocol itself, a few standing measures raise the cost of targeting your family. Freezing credit at the three major bureaus closes off new-account fraud at no cost. Custodian-level protections, including verbal confirmation requirements for withdrawals, trusted-contact designations, and view-only access for aging parents’ accounts, put institutional friction between an impostor and the money. And a measured reduction in the family’s public audio and video footprint, particularly for children and for elderly relatives, shrinks the training data available to a cloning tool. It requires the same attitude toward operational risk that families already apply to insurance, liability, and asset protection.
In Conclusion
The tools behind these schemes will keep improving. What won’t change is the structure of the attack. Someone builds false trust, creates urgency, and asks for money on a channel they control. Families that verify through channels they control, require second approvals, and treat urgency itself as the warning sign take away all three levers at once.
Fraud protection now belongs in the same review as your portfolio, tax, and estate planning, because a single successful scheme can outweigh years of disciplined compounding. If your household hasn’t put a verification protocol in writing, or if aging parents, new liquidity, or a coming transaction has raised the stakes, we’d welcome the chance to help you build one. Schedule a review with our team to get started.
Sources
- FBI Internet Crime Complaint Center, 2025 Annual Report
- Deloitte Center for Financial Services, Generative AI Is Expected to Magnify the Risk of Deepfakes and Other Fraud in Banking
- Federal Trade Commission, Scammers Use AI to Enhance Their Family Emergency Schemes
- CNN Business, Arup Revealed as Victim of $25 Million Deepfake Scam
- FBI Internet Crime Complaint Center, 2024 Annual Report
- Federal Bureau of Investigation, Cryptocurrency and AI Scams Bilk Americans of Billions
