The Prague Post - '20 minutes of terror': AI boosts US voice impersonation scams

EUR -
AED 4.120833
AFN 72.374653
ALL 90.68597
AMD 405.213892
ANG 2.008278
AOA 1028.945251
ARS 1713.128375
AUD 1.597286
AWG 2.021143
AZN 1.906221
BAM 1.95442
BBD 2.261004
BDT 138.851961
BGN 1.961202
BHD 0.423701
BIF 3367.352434
BMD 1.122078
BND 1.434081
BOB 13.162823
BRL 5.641241
BSD 1.122607
BTN 106.103084
BWP 15.379141
BYN 3.334231
BYR 21992.723667
BZD 2.257726
CAD 1.577877
CDF 2598.73214
CHF 0.926595
CLF 0.027838
CLP 1095.630483
CNY 7.493291
CNH 7.490541
COP 3580.190997
CRC 511.139102
CUC 1.122078
CUP 26.941977
CVE 110.187691
CZK 24.274862
DJF 199.898858
DKK 7.474749
DOP 68.941323
DZD 151.101229
EGP 61.024421
ERN 16.831166
ETB 183.87581
FJD 2.52894
FKP 0.838851
GBP 0.83859
GEL 2.90609
GGP 0.838851
GHS 13.083762
GIP 0.838851
GMD 82.471988
GNF 9935.984531
GTQ 8.578943
GYD 234.86417
HKD 8.805101
HNL 30.135722
HRK 7.533407
HTG 147.087457
HUF 365.420883
IDR 20238.130708
ILS 3.465427
IMP 0.838851
INR 105.958979
IQD 1701.798417
IRR 2056067.194777
ISK 136.051813
JEP 0.838851
JMD 178.164204
JOD 0.795528
JPY 177.460525
KES 145.679109
KGS 98.126099
KHR 4564.776962
KMF 493.15641
KPW 1009.869921
KRW 1490.130267
KWD 0.348091
KYD 0.935535
KZT 501.445946
LAK 25267.384721
LBP 100528.116341
LKR 372.695192
LRD 191.404855
LSL 18.415597
LTL 3.313204
LVL 0.678734
LYD 7.202946
MAD 11.018371
MDL 20.195651
MGA 5073.463011
MKD 61.606502
MMK 2355.482879
MNT 4036.712338
MOP 9.073044
MRU 45.027208
MUR 53.063078
MVR 17.347799
MWK 1946.525623
MXN 20.926682
MYR 4.579423
MZN 71.703844
NAD 18.415597
NGN 1494.775718
NIO 41.310832
NOK 10.522281
NPR 169.765134
NZD 1.972383
OMR 0.431441
PAB 1.122617
PEN 3.858176
PGK 5.026518
PHP 71.260352
PKR 310.70134
PLN 4.357926
PYG 6459.439204
QAR 4.09251
RON 5.340639
RSD 117.418663
RUB 91.815241
RWF 1659.121156
SAR 4.206631
SBD 9.060306
SCR 15.520557
SDG 674.931784
SEK 11.108811
SGD 1.434072
SHP 0.838529
SLE 26.840145
SLL 23529.414064
SOS 641.547025
SRD 42.259131
STD 23224.743713
STN 24.482714
SVC 9.823108
SYP 14589.254397
SZL 18.423156
THB 37.711945
TJS 10.339
TMT 3.938493
TND 3.351134
TOP 2.701694
TRY 55.657864
TTD 7.621619
TWD 36.380565
TZS 2967.89529
UAH 50.546611
UGX 4720.666894
USD 1.122078
UYU 45.068179
UZS 13437.512218
VES 987.191539
VND 29013.003036
VUV 133.67989
WST 3.113405
XAF 655.957
XAG 0.019521
XAU 0.000271826755
XCD 3.032471
XCG 2.023172
XDR 0.793367
XOF 655.957
XPF 119.331742
YER 264.929387
ZAR 18.399657
ZMK 10100.046811
ZMW 22.519301
ZWL 361.308574
SSP 6494.085876
MXV 2.362129
  • RYCEF

    0.1700

    17.67

    +0.96%

  • RBGPF

    0.5300

    70.53

    +0.75%

  • CMSC

    -0.1000

    19.53

    -0.51%

  • NGG

    -0.2800

    80.78

    -0.35%

  • VOD

    -0.0300

    15.85

    -0.19%

  • RIO

    -1.2400

    86.92

    -1.43%

  • BTI

    -0.2100

    58.9

    -0.36%

  • RELX

    -0.2600

    39.13

    -0.66%

  • AZN

    -0.3700

    168.45

    -0.22%

  • GSK

    -0.6100

    50.02

    -1.22%

  • BCC

    -0.6900

    75.59

    -0.91%

  • BP

    0.4600

    49.85

    +0.92%

  • BCE

    -0.0900

    19.98

    -0.45%

  • JRI

    -0.0100

    10.62

    -0.09%

  • CMSD

    -0.0800

    19.61

    -0.41%

'20 minutes of terror': AI boosts US voice impersonation scams
'20 minutes of terror': AI boosts US voice impersonation scams / Photo: Chris Delmas - AFP

'20 minutes of terror': AI boosts US voice impersonation scams

Liz Benz still believes the distressed caller's voice was her son's -- the tone, enunciation and cadence all matched her 16-year-old.

Text size:

But it was an AI clone, making the American mother yet another victim of a growing wave of impersonation scams.

Rapidly evolving artificial intelligence technology has demolished the boundaries between reality and fiction, handing cybercriminals strikingly convincing voice cloning tools to steal from people by mimicking loved ones.

Buffalo-based Benz was jolted from her couch by a call from an unknown number. On the line was someone sounding like her son Fred, crying for help.

She was told Fred's friend had been shot and killed, and her son –- who was out at a local football game -– was being held hostage.

The 46-year-old insurance broker and mother of six was instructed to deliver cash to a nearby Walmart to pay off the man holding him.

Eventually, a selfie from Fred smiling at the game returned her to reality: the call, she realized, was an elaborate scam.

"Nothing could have prepared me to hear my son's voice, and nothing could have convinced me that this was a scam until I saw my son with my own eyes," Benz told AFP, her voice trembling.

"It was a good 20 minutes of terror."

- 'Anyone can do it' -

US authorities and consumer advocates are increasingly warning of scams built around impersonating family members.

The FBI said in April that Americans lost over $893 million last year to AI-enabled hoaxes, including voice cloning scams.

Simple internet searches can surface a wide array of voice cloning apps, many available for free, that create realistic AI replicas using small samples -- sometimes only seconds -- of a person's real voice.

"It used to be somewhat hard to make these things. Now anyone can do it in seconds," said Brian Long, chief executive of Adaptive Security, a company offering trainings on AI fraud protection.

"One guy in a room with a keyboard can make an infinite number of attackers," Long told AFP, adding that AI tools can build entire scripts off of snippets of audio captured from social media or voicemail recordings.

Benz's story highlights a familiar script: an emotionally charged call purportedly from a loved one in trouble -- arrested, in a car accident or caught up in a crime -- who needs money.

Then scammers typically pile on pressure, adding voices claiming to be attorneys, courtroom clerks or bank tellers -- a cast of fictitious characters in a chaotic, urgent-sounding call.

- 'Distressed voice' -

Many family-emergency scams do not even require a perfect voice clone.

"A distressed voice saying 'mom, help me' or 'dad, I've been in an accident' may only need to sound believable for a few seconds," Amit Gupta, the vice president of product management at cybersecurity firm Pindrop, told AFP.

"The objective is not perfect voice replication. The objective is creating enough emotional uncertainty and urgency that the victim acts before verifying."

Since taking her story public, Benz said she has received a flood of messages from other victims, many of whom opt to stay anonymous because of the shame attached.

Elderly people are particularly vulnerable, with experts warning of rising cases of "grandparent scams."

The FBI said Americans over 60 reported more than $7.7 billion in losses last year, a significant jump over 2024.

"These are professionals, and when they get people on the phone, they are dealing with amateurs," said Philadelphia attorney Gary Schildhorn, who faced a similar attack in 2020.

Like Benz, he has since partnered with Adaptive Security to raise public awareness about the threat.

In 2023, Schildhorn testified before the US Senate about his experience with a scam call in which a voice impersonating his son Brett claimed he needed to post bail after a drunk-driving arrest.

The call sent Schildhorn, now 73, rushing to his bank.

"When I get to the bank, my phone rings. It's my son," he told AFP. "He's going, 'You've been scammed,'" Schildhorn said.

"I go, 'Brett, I will go to my grave swearing that it was your voice, it was your cadence, it was words you would use. There was no accent. It was you on the phone.'"

T.Kolar--TPP