Cyberpediaby Brain

Executives: targets of whaling and voice cloning

Why executives are targeted by whaling, cloned voices and fake video calls, and the habits that help you avoid paying a fraudulent request.

On this page
  1. 1In 30 seconds
  2. 2What it is
  3. 3Straight answers
  4. 4Inside the attacker's head
  5. 5What you see on screen
  6. 6Under the hood
  7. 7Case files
  8. 8Sources
  9. 9Cite this article

In 30 seconds

Where it happens
In an email presented as coming from a CEO or a chief financial officer[3], in a phone call that imitates an executive's voice[4], or in a video conference where the people on screen are not who they claim to be[2].
What the attacker wants
A payment. In the campaign Microsoft describes, it was an ACH transfer of nearly $50,000 requested from accounting teams[3].
The reflex
Do not answer the request directly. ESET notes that contacting the person by phone, at the official number and not the one in the message signature, can prevent damage[5].
If it is already done
Tell your management and the accounting department, alert the bank as soon as possible to ask for the funds to be recalled, then file a complaint[6].

What it is

An executive decides on payments and opens access. Proofpoint describes whaling as a form of phishing aimed at executives, such as CEOs and chief financial officers, because they have privileged access to sensitive data and financial resources[1].

The attacker may target the executive directly, or take on the executive's identity to push other employees into acting (CEO fraud)[1]. AI tools add voice and face: a cloned voice or a faked face can be inserted into a live call or video conference[2].

Straight answers

Why are executives targeted?

Proofpoint gives two criteria for choosing targets: the authority to approve high-value transactions, and access to sensitive information and systems[1]. According to Proofpoint, executives receive a targeted attempt every 24 days on average[1].

Public exposure also plays a part. Attackers look for information in social media, press releases, public appearances and interviews[1]. Executives with a large public footprint are at higher risk, because a lot of data exists about them[2].

What is the difference between whaling and CEO fraud?

In whaling, the executive is the target. In CEO fraud, the executive is the identity borrowed to get an employee to make a transfer or hand over data[1]. ESET specifies that CEO fraud targets finance or accounting teams[5]. Cybermalveillance.gouv.fr uses the term president scam when the request is presented as coming from an executive and as urgent and confidential[6].

Why do colleagues obey so quickly?

Attackers create a sense of urgency so that the employee asks as few questions as possible[5]. Nobody wants to let down an executive who trusts them, adds ESET[5]. In a call, they may also ask that the request be kept confidential[4].

Does a familiar voice or face prove anything?

No. Proofpoint cites familiarity bias: faced with a known voice or face, people fill in the gaps themselves[2]. Voice clones also copy rhythm, tone and speech mannerisms[2].

What should I do about a pressing transfer request?

For a call, ESET suggests three measures: check the request through another channel (for example the company messaging system), have two people approve large transfers and changes to a supplier's bank details, and agree in advance on passwords or questions that executives answer to prove their identity[4]. Cybermalveillance.gouv.fr recommends an internal verification and hierarchical validation procedure, with no exceptions, for unexpected transfer requests[6].

Am I allowed to slow down or refuse?

Proofpoint describes a culture of "stop-and-verify", in which employees have the knowledge and the permission to stop and think before acting on an unusual request[2].

Do a personal question or an agreed word really help?

Proofpoint reports a CEO fraud attempt aimed at executives: a call reproduced the CEO's voice and accent. It failed when an executive asked a personal question that only the real CEO could know[2].

Inside the attacker's head

ESET describes how a voice attack unfolds. The attacker picks the person to imitate (a CEO, a chief financial officer or a supplier), finds an audio clip online, then picks the person to call, often after research on LinkedIn to find IT support or finance staff[4]. They may send an email before the call, for example an urgent transfer request[4]. According to Proofpoint, three seconds of audio are enough to clone a voice[2].

For an email, Proofpoint describes a first contact that starts with harmless exchanges to build credibility before moving to the financial request[1]. Microsoft describes a campaign that layers executive impersonation, a supplier's branding, a fake invoice and fake exchanges, to reduce the recipient's skepticism[3].

What you see on screen

In an email, Microsoft points out several clues in the campaign it observed:

  • the display name does not match the sender address, and the subject uses financial lure keywords such as "due bill" or "ACH Parment"[3];
  • the "forwarded" thread lacks the headers of a real forward, and its messages are left aligned instead of being indented as in a real thread[3];
  • a phrase such as "no need to copy me" is suspicious[3].

ESET also describes an almost identical address (one letter missing from the domain) or a spoofed address that displays as the legitimate one. In both cases, clicking "Reply" sends the message to the scammer[5]. The following illustration is neutralized and inspired by ESET's example.

Excerpt
Illustration (fictitious addresses in .invalid)
Legitimate address:  [email protected]
Trap address:        [email protected]   [one letter is missing]

On the phone, depending on how sophisticated the AI used is, it may be possible to notice an unnatural speech rhythm, an unusually flat emotional tone, strange or absent breathing, a robotic sound, or background noise that is missing or too uniform[4]. ESET warns that AI flaws may be harder to spot on the phone[4].

Behavioral signals matter just as much, however realistic the image or voice: unusual urgency, a request that skips the usual approval channels, pressure to act outside normal hours[2].

Under the hoodIT teams

According to Microsoft, the attacker registered domains before sending, including a domain imitating ServiceNow registered on July 31, and used several accounts of a third-party email delivery service[3]. Microsoft found no evidence that the legitimate organizations named in the lures, including ServiceNow, were compromised or involved[3]. It noted indicators consistent with the use of generative AI to build the email templates, but they do not establish to what extent AI produced the content[3].

For voices and faces, Proofpoint describes three stages: training on public data about the target, generation, then real-time synthesis, which makes it possible to insert a synthetic voice or face into a call or video conference without prerecorded content[2]. Proofpoint estimates that these fakes do not bypass controls through a technical flaw: they satisfy them[2].

Case filesReal cases

August 3 to 5: more than a million emails imitating executives (spoofed email)

Between August 3 and 5, Microsoft detected a campaign of more than a million emails targeting enterprise users, 87.7% of them in the United States[3]. The attacker posed as CEOs, chief financial officers or presidents of several targeted companies, and asked their accounts payable departments to process an ACH payment of nearly $50,000[3]. The message contained a fake ServiceNow invoice, which ServiceNow did not issue, and a fake exchange between two imitated executives[3].

March 2025: a faked video conference in Singapore

According to Proofpoint, in March 2025 a finance director at a multinational company in Singapore approved a $499,000 transfer after a Zoom call with people who said they were the company's chief financial officer and members of senior leadership[2]. The attackers had suggested the video call, knowing that finance teams were trained to check unusual requests visually[2].

2020: a call supposedly from a director, $35 million

ESET cites, among the heaviest mistakes, a 2020 case: an employee of a company in the United Arab Emirates believed that their director had called to ask for a transfer of $35 million tied to a merger and acquisition operation[4].

2018: two emails to a Canadian city

ESET reports the case of a Canadian city in 2018. Its treasurer transferred more than 100,000 Canadian dollars after a fraudulent email. A second email demanded another 150,000 Canadian dollars. She asked the city manager, whose identity had been spoofed and who was in the same room, whether the request was legitimate, which brought the fraud down[5].

Sources

  1. Proofpoint, “What Is Whaling Phishing? Definition | Proofpoint US”, 5 March 2025. https://www.proofpoint.com/us/threat-reference/whaling (accessed 14 November 2026)
  2. Proofpoint, “What Is Deepfake? Meaning, Technology, How it Works | Proofpoint US”, 26 August 2024. https://www.proofpoint.com/us/threat-reference/deepfake (accessed 14 November 2026)
  3. Microsoft, “Protecting organizations from AI-assisted executive impersonation and invoice fraud”, 10 September 2026. https://www.microsoft.com/en-us/security/blog/2026/09/10/protecting-organizations-ai-assisted-executive-impersonation-invoice-fraud/ (accessed 14 November 2026)
  4. ESET, “Faking it on the phone: How to tell if a voice call is AI or not”, 23 February 2026. https://www.welivesecurity.com/en/business-security/faking-it-phone-how-tell-voice-call-ai/ (accessed 14 November 2026)
  5. ESET, “4:15 p.m.: An urgent message from the CEO”, 13 March 2020. https://www.welivesecurity.com/2020/03/13/415pm-urgent-message-ceo-fraud/ (accessed 14 November 2026)
  6. Cybermalveillance.gouv.fr, “La fraude au virement bancaire (FOVI) ou au faux ordre de virement”, 27 January 2021. https://www.cybermalveillance.gouv.fr/tous-nos-contenus/fiches-reflexes/escroquerie-faux-ordres-virement-fovi (accessed 14 November 2026)