A finance worker at engineering firm Arup fell victim to an elaborate scheme in January 2024 when deepfake technology was used to impersonate company leadership in a video call, resulting in transfers totaling approximately HK$200 million (US$25.6 million). Months later, retail giant Marks & Spencer endured a major breach after attackers deceived third-party helpdesk staff to gain network entry. Most recently, threat actors have posed as IT personnel over Microsoft Teams to extract remote access credentials from employees. Each incident reveals the same vulnerability: organizations struggle to verify that requests for sensitive actions actually come from who they claim to be.

The Arup Deepfake Fraud

In January 2024, an employee in Arup's Hong Kong office received a message purporting to be from the company's UK-based CFO requesting financial transfers. Though initially doubtful, the worker became convinced after joining a video meeting that appeared to show the CFO and other colleagues. The entire conference was fabricated using deepfake technology. According to Hong Kong police, publicly available video and audio materials were weaponized to construct a convincing pre-recorded session. The employee proceeded with multiple transfers, losing approximately HK$200 million.

The M&S Cyberattack

Marks & Spencer experienced a severe cyberattack in April 2025 that disrupted business operations for weeks. The incident suspended online ordering, disrupted contactless payment systems and Click & Collect functionality, and resulted in the theft of customer information. The retailer projected the incident would reduce operating profit for 2025/26 by approximately £300 million before mitigation efforts. The Scattered Spider cybercriminal group, which specializes in social engineering tactics, was identified as responsible. M&S later disclosed that the attackers gained access by manipulating employees at a third-party contractor rather than exploiting a technical flaw in its own infrastructure.

Fake IT Support

Microsoft disclosed a campaign in which attackers posed as IT support personnel using Microsoft Teams to manipulate employees into granting remote device access. The threat actors initiated contact through external Teams messages or calls while impersonating internal IT staff. They employed plausible scenarios—such as claims of urgent technical problems—to persuade targets to enable remote control via screen sharing or established tools like Quick Assist. Following successful connection, the attackers deployed PowerShell to install backdoors, maintain persistent access and extract network intelligence. Microsoft also documented attempts to advance toward sensitive systems, including domain controllers and certificate authorities.

The Common Thread: Trust Replacing Verification

Impersonation Exploits Familiar Trust Signals

Social engineering has long served as a cornerstone of attack strategies. Attackers need not deploy complex malware or technical vulnerabilities if they can convince a target that they represent a trusted entity. Deepfake capabilities amplify this risk by enabling the recreation of recognizable faces and voices. Yet even without artificial intelligence, attackers can construct persuasive scenarios and leverage established communication channels to make fraudulent requests seem routine.

Trusted Processes Can Become Attack Vectors

Password resets and remote support activities are everyday occurrences, yet they present openings for attackers if verification procedures are inadequate. Third-party service providers expand the threat landscape; while an organization may enforce rigorous controls internally, external support teams may lack equivalent identity verification and authorization safeguards.

Employees Can't Be Expected to Spot Every Impersonation Attempt

Security training improves employee awareness of suspicious communications, but even experienced staff can be misled by sophisticated impersonation tactics. Attackers typically conduct reconnaissance to gather details that make their requests appear legitimate and routine. Organizations require dependable verification mechanisms that assist employees during security-critical decisions rather than expecting them to independently assess trustworthiness.

How Can Organizations Mitigate These Risks?

Strengthen Identity Verification at the Service Desk

Service desk teams should move beyond security questions or personal details—which attackers may already possess—and adopt robust authentication methods to confirm caller identities. Solutions like Specops Secure Service Desk enable agents to verify identities using more than 15 MFA factors, including phone-free options, with compatibility for platforms such as Duo and Okta. This approach ensures that sensitive account recovery operations are backed by reliable verification, limiting pathways for attackers to manipulate service desk workflows.

Build Verification Into Business Workflows

As organizations deploy self-service portals, internal chatbots and automated IT systems, these channels must enforce the same verification rigor as traditional support. A password reset initiated through an internal chatbot, for instance, should still demand strong identity verification before processing. Tools like the Specops Authentication API allow organizations to embed secure workforce authentication and identity verification into existing applications, automated processes and business workflows. By weaving verification into these systems, organizations can enforce consistent security checks at critical junctures without depending solely on manual review.

Independently Verify Sensitive Requests

Financial transactions, privileged access grants and other high-risk requests warrant independent confirmation using established channels. This might involve reaching out to the requester through previously confirmed contact information or incorporating additional authentication steps into approval chains.

Verify IT Support Requests Before Granting Access

When support operations run through internal systems or automated workflows, supplementary identity verification can confirm that sensitive requests originate from authorized personnel. These protections should be paired with broader security measures including phishing-resistant MFA and least-privilege access policies, alongside surveillance for unusual account behavior.

Make Strong Verification Part of Everyday Security

As social engineering methods grow more convincing, organizations cannot depend on staff to catch every fraudulent attempt. Robust identity verification, independent approval chains and documented security procedures substantially reduce exposure. By embedding these safeguards into helpdesks, applications and routine workflows, organizations make it substantially harder for attackers to convert a convincing impersonation into an actual breach.