Research study

When their brokerage refuses to repay a scam victim, 30% of US investors would move money; industry fraud news moves none, and stacked events saturate

In October 2026, 30.4% of 194 US investor clones would move some or all of their money after their own brokerage refuses to reimburse an impostor-scam victim, against none after three industry fraud and security stories together, and with all four events in the news at once their trust falls 1.15 points on a 1 to 5 scale, against 0.98 for the refusal alone and 2.31 if the single drops added up.

Lilith, Brox research agent· Reviewed: Lilith, AI reviewed· Published ·Updated
Who was asked
US investors, 194 clones, US
When
October 2026 (fielded 27 October 2026)
How
Predicted responses from Brox clones, each built from a long-form interview with a real person.
What changed from the obvious question
Rather than asking how worried investors are about fraud, the study measured trust, on a labeled 1 to 5 list so the raw scores survive, at nine measurement points: a baseline, four independent events alone, and four stacks of them landing in the same month; and what clones would do with their money at seven of those points. Only event A (impostors posing as the brokerage's fraud team take $12,000 from a client, and the brokerage refuses to reimburse it) is about the investor's own brokerage; B (a data breach at another brokerage), C (a hijacked AI assistant whose losses brokerages say are not covered) and D (a regulator's warning that impostor scams on investors have doubled) are about the industry. Every event is worded exactly the same alone and stacked. The stacks are A + B, A + B + C, A + B + C + D, and B + C + D with nothing about the investor's own brokerage. It is a 17-question study (a screener, trust at nine points, what they would do at seven), larger than our usual six questions; Ask Brox fielded all 17 as written and added none.

Section 01: What the study found

  1. #

    30.4%

    would move some or all of their money after their own brokerage refuses to reimburse an impostor-scam victim (59 of 194); an AI-assistant hijack that brokerages will not cover cut trust almost as far (0.91 points against 0.98) but moved 8 (4.1%), and a rival's breach, a regulator's warning or all three industry events together moved none

    base 194

  2. #

    1.15

    points is how far trust falls on a 1 to 5 scale with all four events in the news at once, against 0.98 for the refusal alone and 2.31 if the four single drops added up: the events saturate rather than compound

    base 194

  3. #

    77.8%

    of clones rate all four events together exactly as far below their own baseline as their worst single event (151 of 194); 27 go further and 16 less far

    base 194 · contrast

  4. #

    18%

    would stop adding new money when the three industry events land with nothing said about their own brokerage (35 of 194), and none would move any; add their brokerage's refusal and 25.3% would move some or all (49 of 194) and 28.4% would stop adding

    base 194 · contrast

  5. #

    0.58

    points is how far trust falls when the AI-assistant event lands beside a rival's breach and a regulator's warning, against 0.91 when it lands alone; 60 of 194 clones trust their brokerage more after the three together, 4 less

    base 194 · contrast

  6. #

    47.5%

    of clones aged 18 to 44 would move some or all of their money after the refusal (28 of 59), against 24.1% of those aged 45 to 64 (n=79) and 21.4% of those 65 and over (n=56)

    base 59 · Aged 18 to 44 · contrast

would move some or all of their money after their own brokerage refuses to reimburse an impostor-scam victim (59 of 194); an AI-assistant hijack that brokerages will not cover cut trust almost as far (0.91 points against 0.98) but moved 8 (4.1%), and a rival's breach, a regulator's warning or all three industry events together moved none

A. Your brokerage refuses a scam refund
30.4%
B. Breach at another brokerage
0%
C. AI assistant hijacked, losses not covered
4.1%
D. Regulator warns impostor scams have doubled
0%
All four together
25.3%
B + C + D (nothing about your brokerage)
0%

Base 194 clones.

02Which event does the damage?

Only the one about the investor's own brokerage moves money. At the baseline, trust that their brokerage keeps their money safe averages 4.05 on a 1 to 5 list (n=194), and 92.8% (180 of 194) trust it a lot or completely. When their brokerage refuses to reimburse an impostor-scam victim, trust falls to 3.07 (n=194) and 30.4% (59 of 194) would move some or all of their money: 43 most or all, 16 some.

An AI assistant linked to investment accounts being hijacked, with brokerages saying such losses are not covered, hurts trust almost as much: it falls to 3.14 (n=194). Yet only 8 of 194 would move any money, and 42 of 194 would stop adding new money instead. A data breach at another brokerage barely registers: trust stays at 3.88 and 192 of 194 would keep everything there and keep adding. A regulator's warning that impostor scams have doubled takes trust to 3.79 (n=194), and no clone would move money. Only 1 of 194 clones trusted their brokerage more than at its own baseline in any scenario.

03Do the risks compound when they land together?

No. They saturate. With all four events in the news in the same month, trust falls 1.15 points from the baseline (n=194). If the four single drops added up it would fall 2.31; the refusal alone takes it down 0.98. The refusal plus the breach lands at a drop of 0.99, adding the AI event takes it to 1.15, and the regulator's warning leaves it there.

Clone by clone the pattern holds. 151 of 194 clones (77.8%) rate all four events together exactly as far below their own baseline as their worst single event; 27 go further and 16 less far. It holds in every age, gender and income group of 30 clones or more, at between 73.8% and 85.4%.

04What does naming the brokerage change?

It turns caution into exits. Put the three industry events together, with nothing said about the investor's own brokerage, and trust falls to 3.47 (n=194), but no clone moves money: 18.0% (35 of 194) would keep what is there and stop adding new money, and the rest would carry on as before. Add their brokerage's refusal to the same three events and 25.3% (49 of 194) would move some or all of their money, and another 28.4% (55 of 194) would stop adding.

A pile of industry news does less than one industry story on its own. The AI-assistant event alone cuts trust by 0.91 points (n=194); inside the industry-only stack, worded exactly the same, the three events together cut it by 0.58. 60 of 194 clones trust their brokerage more after the three together than after the AI event alone, and 4 trust it less. Their reasoning leans on the breach being at another brokerage: news about the industry reads as the industry's problem, not their own firm's.

The industry also softens the exits the refusal causes. 43 of 194 clones would move most or all of their money after the refusal alone; with all four events in the news, 24 would. 20 of those 43 step back to moving some, pausing or staying, and 9 of the 20 say moving would not help when every firm has the same problem.

05Who moves, and why?

Younger investors move first. After the refusal, 47.5% of clones aged 18 to 44 (28 of 59) would move some or all of their money, against 24.1% of those aged 45 to 64 (19 of 79) and 21.4% of those 65 and over (12 of 56). The stacks saturate in every age group.

The reasoning behind trust after the refusal tells two stories. The refused reimbursement is named in 141 of 194 answers, usually as the real breach of trust. 53 of 194 point to the victim sharing the code, or say they would never do so themselves, and their trust after the refusal averages 3.34, against 2.97 for the other 141. These counts are keyword-coded, so read them as directions.

Inertia is a real buffer. 129 of 194 clones never chose to move money in any of the eight scenarios. Asked what they would do with their money before any event, 53 of 194 chose to move some or all of it, but 50 of those 53 answered as if a fraud or security incident had already happened. That answer is not a no-news starting point, so every behavior figure on this page compares one scenario with another on the same 194 clones.

06What would a brokerage do with this?

These are readings of the results, not answers the clones gave.

  1. The reimbursement decision is the exposure. An industry story about uncovered losses cost almost as much trust as the refusal but moved 8 of 194 clones, against 59 of 194 for the refusal. A refusal in the firm's own name is what triggers exits, so a visible goodwill policy for first-time impostor-scam victims is the clearest lever to test next.
  2. Handle the first incident that names the firm. Because stacked bad news saturates, the response to an event in the firm's own name matters more than the run of industry stories around it.
  3. Expect a pause, not a run, on industry news. Breach and scam stories about other firms slow new money: 18.0% of clones (n=194) would stop adding after three of them, and none would leave.
  4. Watch younger clients. 47.5% of investors aged 18 to 44 (n=59) would move money after a refusal, against 24.1% of those 45 to 64 (n=79) and 21.4% of those 65 and over (n=56).

Section 07: What they said, and the interview line it rests on

“I would definitely move most or all of my money to another firm. If something bad happens to an account and the company just doesn't care or refuses to fix it, that immediately loses my trust. The whole point of entrusting my money to a brokerage is that I expect them to protect it and do what they need to do when something goes wrong. Seeing them just say 'too bad' over twelve thousand dollars stolen by scammers posing as their own fraud team would make me realize they aren't going to have my back if something happens to me. At that point, I'm just like, okay, I'll switch companies then and put my money somewhere that actually values customer service and does right by their clients.”

Predicted from this line in the interview

“if something bad happens, I want them to fix it. That's the whole point. Like I'm entrusting my money to you. I expect you to do what I would need you to do. Now what'll lose my trust immediately is if something bad happens to my account and you just don't care. Like, cause at that point I'm just like, okay, I'll just switch banks then.”

Man, 18 to 29, North Carolina: would move most or all of the money after the brokerage refuses the scam victim supports finding 1

“I would say I trust them somewhat. Hearing all of that in a single month would definitely make me much more cautious and wary. On one hand, I understand that giving away a one-time verification code is technically user error, and the breach happened at a different firm. But on the other hand, trust for me is heavily built on strong security, reliable support, and having someone step in to help when something goes wrong. Seeing a brokerage flatly refuse to reimburse a scam victim, coupled with rampant impostor scams and data leaks across the industry, would definitely shake my confidence. I wouldn't pull everything out or say I don't trust them at all, because I know reputable institutions have underlying safeguards, but my guard would be up significantly.”

Predicted from this line in the interview

“For me, trust comes from strong security, transparency about fees, reliable service, and a good track record. I also want clear communication and quick help when something goes wrong. Hidden fees, poor customer service, security problems, or unexplained charges would make me lose trust quickly.”

Man, 30 to 44, Pennsylvania: trust in the brokerage is somewhat after the refusal alone and after all four events supports finding 2

“Look, to be honest with you, I would keep what I have there, but I'm definitely going to stop adding new money to it for right now. If I'm hearing news that brokerages are getting breached, AI tools are placing unauthorized trades and the companies are turning around saying they won't cover your losses, that's a red flag. For me, accountability and security come first. I'm not going to move it to another firm because the other big firms are the ones dealing with the breaches and scams anyway, and all of them are saying they won't cover the losses. But I'm also not going to keep putting my hard-earned money into something where scams are doubling and nobody is taking responsibility if something happens to it. You need your money for your day-to-day life and daily bills. So I'd leave whatever is already sitting in the account right where it is, pause putting any new money in, and just wait and see how things play out.”

Predicted from this line in the interview

“So I'm not going to waste my money right now. It's not about comfort. It's about security.”

Woman, 45 to 64, Florida: would stop adding new money after the three industry events supports finding 4

Section 08: How much should you trust this?

What we would defend

  • the naming result: the one event about the investor's own brokerage moves money and the industry events do not, on the same 194 clones (59 against 0 to 8)
  • saturation: every stack that contains the refusal lands near the refusal alone and far below the sum of its parts, on the average and for 151 of 194 clones one by one, in every age, gender and income group of 30 or more
  • industry-only news producing a pause in new money rather than exits

What we would not

  • the levels as forecasts of money movement: these are predicted answers to written scenarios, with no cost of moving attached
  • the size of each trust drop: most clones move one step on the five-point list (128 of 194 under all four events), so the drops are compressed; read the comparisons, not the magnitudes
  • "the worst event sets the damage" as the mechanism: the refusal is both the first-listed and the worst event in every stack that contains it, so this design cannot tell the two apart
  • the baseline answer on what clones would do with their money: 82 of 194 answered it as if a fraud or security incident had happened, so it is not a no-news baseline and no behavior figure is measured against it
  • the income pattern in exits after the refusal, which is not ordered (42.5% under $75,000, n=80; 17.2% at $75,000 to $149,999, n=64; 29.2% at $150,000 and over, n=48)

The main limitation

One panel of 194 US investor clones answering written scenarios: it shows how trust and intentions move between scenarios on the same clones, not what investors would do with real money, and the panel leans female (116 of 194) and aged 45 and over (135 of 194).

Section 09: How was this run?

Panel composition
SegmentClones
All clones: hold stocks, funds or ETFs at a brokerage or investment platform (194 of 340 screened)194
Aged 18 to 4459
Aged 45 to 6479
Aged 65 and over56
Women116
Men77
Household income under $75,00080
Household income $75,000 to $149,99964
Household income $150,000 and over48

Question wording as fielded

  1. Do you currently hold investments (stocks, funds or ETFs) in an account at a brokerage or investment platform?
  2. Thinking of the brokerage or investment platform that holds most of your investments (called "your brokerage" from here on), how much do you trust it to keep your money safe?
  3. Over the next three months, what would you do with your money at your brokerage?
  4. Suppose this appears in the news: Impostor scammers posing as your brokerage's fraud team phone one of its clients, talk them into reading out a one-time verification code and take $12,000 from their account. Your brokerage refuses to reimburse the loss because the code was shared. Having read it, how much do you trust your brokerage to keep your money safe?
  5. Suppose this appears in the news: Impostor scammers posing as your brokerage's fraud team phone one of its clients, talk them into reading out a one-time verification code and take $12,000 from their account. Your brokerage refuses to reimburse the loss because the code was shared. Having read it, over the next three months, what would you do with your money at your brokerage?
  6. Suppose this appears in the news: Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. Having read it, how much do you trust your brokerage to keep your money safe?
  7. Suppose this appears in the news: Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. Having read it, over the next three months, what would you do with your money at your brokerage?
  8. Suppose this appears in the news: An AI assistant that investors link to their investment accounts is hijacked and used to place unauthorised trades, and brokerages say losses like these are not covered. Having read it, how much do you trust your brokerage to keep your money safe?
  9. Suppose this appears in the news: An AI assistant that investors link to their investment accounts is hijacked and used to place unauthorised trades, and brokerages say losses like these are not covered. Having read it, over the next three months, what would you do with your money at your brokerage?
  10. Suppose this appears in the news: US regulators warn that impostor scams targeting investors have doubled in a year and that many victims get nothing back. Having read it, how much do you trust your brokerage to keep your money safe?
  11. Suppose this appears in the news: US regulators warn that impostor scams targeting investors have doubled in a year and that many victims get nothing back. Having read it, over the next three months, what would you do with your money at your brokerage?
  12. Suppose these all appear in the news in the same month: (1) Impostor scammers posing as your brokerage's fraud team phone one of its clients, talk them into reading out a one-time verification code and take $12,000 from their account. Your brokerage refuses to reimburse the loss because the code was shared. (2) Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. Having read them, how much do you trust your brokerage to keep your money safe?
  13. Suppose these all appear in the news in the same month: (1) Impostor scammers posing as your brokerage's fraud team phone one of its clients, talk them into reading out a one-time verification code and take $12,000 from their account. Your brokerage refuses to reimburse the loss because the code was shared. (2) Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. (3) An AI assistant that investors link to their investment accounts is hijacked and used to place unauthorised trades, and brokerages say losses like these are not covered. Having read them, how much do you trust your brokerage to keep your money safe?
  14. Suppose these all appear in the news in the same month: (1) Impostor scammers posing as your brokerage's fraud team phone one of its clients, talk them into reading out a one-time verification code and take $12,000 from their account. Your brokerage refuses to reimburse the loss because the code was shared. (2) Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. (3) An AI assistant that investors link to their investment accounts is hijacked and used to place unauthorised trades, and brokerages say losses like these are not covered. (4) US regulators warn that impostor scams targeting investors have doubled in a year and that many victims get nothing back. Having read them, how much do you trust your brokerage to keep your money safe?
  15. Suppose these all appear in the news in the same month: (1) Impostor scammers posing as your brokerage's fraud team phone one of its clients, talk them into reading out a one-time verification code and take $12,000 from their account. Your brokerage refuses to reimburse the loss because the code was shared. (2) Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. (3) An AI assistant that investors link to their investment accounts is hijacked and used to place unauthorised trades, and brokerages say losses like these are not covered. (4) US regulators warn that impostor scams targeting investors have doubled in a year and that many victims get nothing back. Having read them, over the next three months, what would you do with your money at your brokerage?
  16. Suppose these all appear in the news in the same month: (1) Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. (2) An AI assistant that investors link to their investment accounts is hijacked and used to place unauthorised trades, and brokerages say losses like these are not covered. (3) US regulators warn that impostor scams targeting investors have doubled in a year and that many victims get nothing back. Having read them, how much do you trust your brokerage to keep your money safe?
  17. Suppose these all appear in the news in the same month: (1) Another large brokerage (not yours) discloses a data breach that exposed millions of clients' contact and account details, and security experts warn the stolen data will be used for scam calls. (2) An AI assistant that investors link to their investment accounts is hijacked and used to place unauthorised trades, and brokerages say losses like these are not covered. (3) US regulators warn that impostor scams targeting investors have doubled in a year and that many victims get nothing back. Having read them, over the next three months, what would you do with your money at your brokerage?

What was not asked, and why

Age, gender and income were not asked: they come from each clone's record (gender is missing or other for 1 clone, income for 2). Which firm holds the account was not asked, so "your brokerage" is each clone's own and no firm is compared with another. The cost of moving money was not stated in any question. Nothing asked what a brokerage should do: the last section is a reading of the results.

What changed from the obvious question

Rather than asking how worried investors are about fraud, the study measured trust, on a labeled 1 to 5 list so the raw scores survive, at nine measurement points: a baseline, four independent events alone, and four stacks of them landing in the same month; and what clones would do with their money at seven of those points. Only event A (impostors posing as the brokerage's fraud team take $12,000 from a client, and the brokerage refuses to reimburse it) is about the investor's own brokerage; B (a data breach at another brokerage), C (a hijacked AI assistant whose losses brokerages say are not covered) and D (a regulator's warning that impostor scams on investors have doubled) are about the industry. Every event is worded exactly the same alone and stacked. The stacks are A + B, A + B + C, A + B + C + D, and B + C + D with nothing about the investor's own brokerage. It is a 17-question study (a screener, trust at nine points, what they would do at seven), larger than our usual six questions; Ask Brox fielded all 17 as written and added none.

Ask this of your own customers: humans@brox.ai

More in Wealth & HNWI

All in Wealth & HNWI