THREAT ASSESSMENT REPORT • JANUARY 2026

The Invisible

Workforce

How North Korean state actors, deepfake technology, and organized interview farms are embedding operatives inside European and global companies right now.

"The threat of unintentionally hiring North Korean IT workers is larger than most people realize. It is covert, it is global, and it is active right now."

— Kevin Mandia, Former CEO of Mandiant
0%
Increase in infiltrations
0+
Fraudulent hires detected (Okta)
0+
Companies targeted globally
$0M
Illicit revenue, single operation
Scroll
◼ THREAT SCOPE

The Numbers Don't Lie

What began as isolated incidents has become an industrial-scale global operation. These are the documented numbers — and they represent only what was caught.

1
per day
North Korean IT worker incident investigated by CrowdStrike
85
healthcare
Fraudulent interview attempts on healthcare orgs in 2025 alone
27%
non-US
Of identified fraudulent interviews now target non-US companies
70 min
to create
Time needed to create a convincing deepfake — no expertise required
GARTNER PREDICTION • JULY 2025
1 in 4

Candidate profiles will be fake by 2028.

Your current hiring process was not designed for this. Nearly every Fortune 500 CISO has already encountered this threat inside their organization.

European Infiltration Map

LIVE
Country
Primary Sector Targeted
Status
Threat Level
Germany
Energy Infrastructure
Confirmed
CRITICAL
United Kingdom
Healthcare / NHS
Confirmed
CRITICAL
Poland
Financial Services
Confirmed
CRITICAL
Czech Republic
Cloud Technology
Confirmed
HIGH
Romania
Laptop Farm Hub
Active
HIGH
Austria
Critical Infrastructure
Confirmed
HIGH
Portugal
Multiple Sectors
Reported
MEDIUM

* CrowdStrike reports investigating European incidents daily. Sources: FBI, ENISA, CrowdStrike, Mandiant, Palo Alto Unit 42.

◼ ATTACK ANATOMY

How the Attack Works

From a synthetic LinkedIn profile to an operative with admin access to your production systems — this is the five-stage playbook.

01
👤
Identity Construction
Weeks before application

AI generates a complete synthetic persona: name, photo (ThisPersonDoesNotExist.com), LinkedIn profile, fabricated work history, AI-generated references, and identity documents using Midjourney. The persona is optimized for the target region — Western names, European addresses, convincing digital footprints.

Tools used:MidjourneyChatGPTThisPersonDoesNotExist.comAI doc generators
02
📄
AI-Optimized Application
Application stage
03
🎭
Interview Farm
Interview stage — Asia
04
💻
Placement & Laptop Farm
First 30–90 days
05
⚠️
Exfiltration & Leverage
Ongoing — often months
🔑

Your intrusion detection system won't trigger — because there's no intrusion.

The operative has valid credentials, approved system access, and a legitimate employment contract. Traditional security doesn't catch this. The human firewall does.

◼ DOCUMENTED INCIDENTS

Case File Dossiers

Eleven documented incidents. Across multiple sectors. On three continents. All within 24 months. Click any case to open the file.

🇺🇸
US-2026-DLV
Mar 2026
CRITICAL

Delve — Fake Compliance-as-a-Service

📍 United States🏭 Compliance Services
Hundreds of client compliance certifications invalidated; Delve halted operations; Insight Partners distanced itself; regulatory fallout for clients ongoing

Delve, a Y Combinator-backed compliance startup ($32M Series A at a $300M valuation, led by Insight Partners), was accused of fabricating compliance certifications for hundreds of customers — a practice whistleblower 'DeepDelver' called 'structural fraud.' Delve generated auditor conclusions, test procedures, and final report evidence before any independent review occurred, then had partner audit firms (Accorp and Gradient — both nominally India-based, with only a nominal US presence) rubber-stamp the results. Customers received certifications claiming HIPAA, SOC 2, and GDPR compliance they had never actually achieved — exposing them to criminal regulatory liability and significant fines. 'By generating auditor conclusions before independent review, Delve places itself in the role of both implementer and examiner. This is not a technicality. It is structural fraud that invalidates the entire attestation,' wrote DeepDelver. In addition, a security researcher separately reported accessing Delve's own sensitive data — including employee background checks and equity vesting schedules — through multiple 'gaping security holes' in its external attack surface. Delve halted all demos; Insight Partners deleted its investment blog post. Delve denies wrongdoing, calling the pre-filled evidence 'templates.' The direct relevance to insider threat programs: when onboarding IT vendors or staffing providers, organizations routinely accept compliance badges and background check certificates at face value. Delve proves such paperwork can be fabricated at industrial scale. Require the actual audit trail — not just the certificate — and always verify independently.

▶ Open case file
🇨🇿
CZ-2026-PAR
2026
HIGH

LPP Holding — Pardubice Arson Attack

📍 Czech Republic🏭 Defense Industry
One warehouse destroyed; 3 arrests; stolen documents held as extortion leverage

On March 20, 2026, a warehouse at LPP Holding's Pardubice industrial facility was deliberately set ablaze. 'The Earthquake Faction' claimed responsibility, citing the company's alleged cooperation with Israeli drone manufacturer Elbit Systems. One warehouse destroyed; second building damaged; no casualties. The Ministry of Interior classified the event as a probable terrorist attack; BIS (Czech Security Information Service) joined the investigation. The operational precision of the attack raises questions about possible insider intelligence — shift schedules, camera positions, guard routines. UPDATE (March 24): The attack had a second dimension — the group also exfiltrated confidential documents from the facility during the assault. Days later, The Earthquake Faction issued an extortion letter demanding LPP Holding publicly distance itself from Elbit Systems and condemn the 'occupation of Palestine' by April 20 — or the stolen documents will be published. LPP maintains the Elbit collaboration was planned two years ago but never materialized. Three suspects were arrested (one in Slovakia, two in Czech Republic; US and Czech nationals). Intelligence analysts warn the 'pro-Palestinian' framing may be a deliberate smokescreen: LPP Holding supplies drones to Ukraine, giving Russia direct motive to disrupt the factory. 'The language is written to evoke left-wing activism — but you cannot draw conclusions from the text alone,' said analyst Josef Kraus. Whether this is genuine activism or a state-sponsored false flag operation remains unresolved.

▶ Open case file
🇨🇿
CZ-2026-SLP
2026
HIGH

Slavia Pojišťovna — Czech Insurance Provider

📍 Czech Republic🏭 Insurance
Under active investigation

In March 2026, Czech insurance company Slavia Pojišťovna allegedly suffered a 150GB data exfiltration. Leaked data includes medical records, ultrasound imagery, insurance policies, accident and claims reports, and a customer database. The attacker exploited an Adminer vulnerability to deploy a webshell and reverse shell for persistent extraction. Critically, the threat actor claims active network access is still maintained at time of reporting. Investigation is ongoing — whether an insider facilitated initial access or lateral movement remains unconfirmed. This is a live case study in how external breach and insider threat converge.

▶ Open case file
🇩🇪
DE-2026-BER
Jan 2026
CRITICAL

Berlin Power Grid — Vulkangruppe Arson

📍 Germany🏭 Critical Infrastructure / Energy
~100,000 people affected; estimated €tens of millions in emergency response and grid repair

In the early hours of January 4, 2026, far-left militant group Vulkangruppe (Volcano Group) set fire to high-voltage cables on a bridge over the Teltow canal near the Lichterfelde power station in south-west Berlin. The attack cut electricity to 45,000 households and 2,200 businesses across four districts — the longest blackout in Berlin's post-war history. Power was not fully restored until January 8, leaving ~100,000 people without heat for five days in below-freezing temperatures. Schools, hospitals, and care homes ran on emergency generators. In a 2,500-word statement, the group cited the climate crisis and AI data center energy consumption as its motivation — explicitly targeting 'the ruling class.' Initial authorities considered Russian state sabotage, underscoring the operational sophistication of the strike. The attack was not the group's first: Vulkangruppe previously claimed a 2024 arson at the Tesla gigafactory pylon and a September 2025 attack that blacked out south-east Berlin for 60 hours. The precision of target selection — disabling the specific cables with maximum grid impact — raises the question of whether reconnaissance benefited from insider knowledge of maintenance windows, patrol schedules, or cable routing. This remains uninvestigated.

▶ Open case file
🇬🇧
UK-2025
2025
CRITICAL

NHS Synnovis Data Breach

📍 United Kingdom🏭 Healthcare
£22M+ in response costs

A database administrator hired by NHS pathology contractor Synnovis — serving major London hospitals — spent six months systematically exfiltrating 4.2 million patient records: medical histories, mental health assessments, genetic test results. The breach directly contributed to a ransomware attack that disrupted services across multiple NHS facilities. The operative claimed British citizenship with flawless documentation.

▶ Open case file
🇺🇸
US-2024-FARM
2024–2025
CRITICAL

Chapman Laptop Farm Operation

📍 United States🏭 Multi-sector
$17.1M in fraudulent salary

Christina Chapman, 50, operated 90 company-issued laptops at her Arizona home — all received on behalf of North Korean workers. She installed remote access software so operatives overseas could work as if in the US. Result: 309 fraudulent hires across major US companies, $17.1M in illicit revenue. Nearly 70 Americans had their identities stolen. Nike confirmed it was among the victims. Chapman received an 8.5-year prison sentence.

▶ Open case file
🇪🇺
EU-2024-INT
2024–2025
HIGH

European Interview Farm Operations

📍 Multiple EU Countries🏭 IT Recruitment
Ongoing — scale unknown

Czech IT recruiters documented a consistent pattern: perfect resumes, flawless written communication, then on video calls — cameras 'malfunctioning', multiple voices in background, IP addresses showing Asian locations despite EU residence claims. When confronted, candidates immediately disconnected and never responded again. In 9 of 10 such cases, the candidate vanished. CrowdStrike has since confirmed laptop farms in Romania and Poland.

▶ Open case file
🇩🇪
DE-2024
2024
CRITICAL

German Energy Infrastructure Operator

📍 Germany🏭 Critical Infrastructure / Energy
Classified — regulatory fines included

A major German energy operator (classified under EU CER Directive and KRITIS) discovered an IT contractor had gained privileged access to operational technology systems. The contractor, claiming credentials from Eastern European firms, was later identified as a fraudulent operative. The breach triggered mandatory incident reporting under NIS2 and German KRITIS regulations, and exposed vulnerabilities in third-party contractor screening policies.

▶ Open case file
🇵🇱
PL-2024
2024
HIGH

Warsaw Financial Institution

📍 Poland🏭 Financial Services
€840,000 diverted

A senior developer at a Warsaw financial firm worked for 8 months before his scheme was uncovered: he had embedded code that rounded down wire transfers by fractions of a cent, diverting the difference to cryptocurrency wallets. Over 8 months, €840,000 was siphoned. His Polish citizenship documents and claimed employment history at Estonian and Lithuanian banks were entirely AI-fabricated. Polish regulators used the case to push for enhanced personnel security in finance.

▶ Open case file
🇺🇸
US-2024-KB4
2024
CRITICAL

KnowBe4 — A Cybersecurity Company

📍 United States🏭 Cybersecurity
Contained

KnowBe4 — a company that trains organizations on cybersecurity — inadvertently hired a North Korean IT worker. He passed background checks, conducted professional interviews, and received a company laptop. Within minutes of receiving it, he attempted to load malware onto the corporate workstation. Security systems detected the suspicious activity before damage occurred. The incident became a landmark case study in the threat.

▶ Open case file
🇨🇿
CZ-2023
2023
HIGH

Czech Cloud Services Company

📍 Czech Republic🏭 Cloud Technology
Avoided

A software developer named 'Denys Emil L.' claimed Danish citizenship and applied via LinkedIn. Documents looked professional — but Scaut's analysts found they were AI-generated fakes. He had borrowed a real Danish citizen's identity and trade license. Investigation revealed a Chinese national systematically targeting multiple Western tech companies simultaneously, later linked to a North Korean-affiliated group.

▶ Open case file
◼ INTERACTIVE EXERCISE

Spot the Fake

Below is a real candidate profile that came through a European tech company's hiring pipeline. Click on every element you think is a red flag. Can you catch what their HR team missed?

Red flags found: 0 / 7
🧑‍💼
Michael Chen
Senior Cloud Infrastructure Engineer
📍 Berlin, Germany 🇩🇪
LinkedIn Member Since
2 months ago
Connections
12 connections
click
Current Employer
Scout Law Tech GmbH
click
Education
Humboldt-Technische Universität Berlin — MSc Computer Science
click
Contact Number
+1 (415) 555-0192 — WhatsApp only
click
Experience
8+ years in cloud infrastructure and DevSecOps
click
Payment Preference
Contractor via Ruul.io or direct crypto preferred
click

Below is a real candidate profile that came through a European tech company's hiring pipeline.

◼ ORGANIZATIONAL ASSESSMENT

Is Your Organization at Risk?

8 questions. No login required. Designed for HR leaders, CISOs, and security teams. Find out where your hiring process is most vulnerable.

Progress0/8
01Identity Verification

When you hire someone, do you verify their identity documents independently (not just accept copies)?

02Employment History

How do you verify a candidate's previous employment?

03Video Interviews

For remote hires, how do you conduct video interviews?

04Contractor Screening

Do you screen contractors and third-party service providers to the same standard as employees?

05Cross-Border Awareness

For candidates claiming non-local residence (EU but from abroad), do you verify foreign criminal records?

06HR–Security Coordination

Is there clear ownership between HR and cybersecurity teams for employment fraud prevention?

07Deepfake Awareness

Are your recruiters trained to detect deepfakes and interview-farm red flags?

08Continuous Monitoring

After hiring, do you continuously monitor privileged employees for behavioral anomalies?

Answer all 8 questions to continue

◼ INTELLIGENCE BRIEFING

Listen & Watch

Explore the threat in depth — through an AI-generated podcast and an upcoming video briefing.

NOTEBOOKLM PODCAST

Deepfake Workers Fund North Korean Nukes

An AI-generated deep-dive into how North Korean IT workers are systematically funding weapons programs through employment fraud — and what organizations can do right now.

0:00~20 min
VIDEO BRIEFING

The Invisible Workforce — Video Briefing

A visual walkthrough of the full threat report — covering the infiltration tactics, real-world cases, and what your organisation can do right now.

◼ DETECTION PLAYBOOK

The Human Firewall

Background screening is your first and most critical line of defense. Here's what to look for — and what to do.

🎥Video Interview Red Flags
Camera frequently 'malfunctioning' or disabled throughout
Multiple voices audible in background (call center environment)
Immediate disconnection when asked about location or surroundings
Answers to technical questions flow perfectly, personal questions cause hesitation
Lighting inconsistencies or facial movements that look unnatural
Ask candidate to wave hand in front of face — deepfakes glitch

Strategic Recommendations

IMMEDIATE

Assign clear ownership

Create a cross-functional team (HR + IT Security + Legal). Employment fraud prevention must have an owner — not fall between departments.

IMMEDIATE

Mandate video with verification

Require video for all remote interviews. Ask candidates to describe their surroundings. Request environmental checks. Record interviews for review.

HIGH

Deploy document authentication

Use Regula Forensics, Trustmatic, or Scaut to verify identity documents. AI-generated IDs are indistinguishable to the human eye.

HIGH

Extend screening to contractors

Fraudulent candidates are found 3–4× more often in contractor pipelines. Make background verification a contractual obligation for all third parties with system access.

STANDARD

Train recruiters on deepfakes

Specific deepfake detection techniques: hand wave test, window description, audio desync. FBI guidance is publicly available — distribute it.

STANDARD

Implement continuous monitoring

Pre-employment screening is not enough. Monitor privileged users for behavioral anomalies post-hire. The NHS breach ran 6 months before detection.

YOUR FIRST LINE OF DEFENSE IS YOUR HIRING PROCESS

The threat doesn't break in.
It gets invited.

Scaut is Europe's first provider of automated background screening purpose-built for the modern threat landscape. Founded in Prague in 2020, we've been tracking DPRK employment fraud since its earliest days — and building the tools to stop it.

Visit Scaut.com →

Sources: FBI, ENISA, CrowdStrike, Mandiant / Google Threat Intelligence Group, Palo Alto Unit 42, DTEX Systems, Okta, Gartner. Report compiled January 2026.

© 2026 Scaut.com — Advanced Background Screening, Prague, Czech Republic