AI

Ok, it’s late, and well I wanted to look into cyber attacks where social engineering is a key component combined with technical hacking skills.

There’s been a growing number of these style events, so I tasked GROK to create an assessment for me, let’s see how it did! Let’s both try and answer the questions:

Can GROK replace intelligence officers and can GROK help us defend better against social engineering + technical attacks? What do you think? (please take all of this with a pinch of salt… LLMs are known to make mistakes/hallucinate/lie in a very convincing manner)….

they look nice…. but looks can as we know, be deceiving! (is the entire blog just a social engineering experiment by me?)

Intelligence Assessment: Scattered Spider and DragonForce Ransomware Threats

Intelligence Assessment: Scattered Spider and DragonForce Ransomware Threats

Date: 28 April 2025

Prepared by: [Intelligence Officer, Name Classified]

Classification: Restricted

Executive Summary

Scattered Spider and DragonForce are distinct cybercriminal entities with sophisticated capabilities to execute ransomware attacks against organizations. Scattered Spider, a Western-based group known for social engineering and cloud exploitation, has a history of targeting large corporations, often collaborating with ransomware-as-a-service (RaaS) providers like ALPHV/BlackCat, RansomHub, and Qilin. DragonForce, a lesser-known but emerging threat actor, is a ransomware group with suspected ties to politically motivated operations, focusing on high-impact targets. Both groups leverage advanced tactics, techniques, and procedures (TTPs) to infiltrate, persist, and extort organizations, posing significant risks to critical infrastructure, financial, and technology sectors. This assessment outlines their methodologies, potential targets, and mitigation strategies.

1. Scattered Spider Overview

Profile:

  • Aliases: UNC3944, Octo Tempest, Muddled Libra, Starfraud, Scatter Swine, 0ktapus.
  • Origin: Predominantly U.S. and U.K.-based, comprising young (19–22 years old) native English-speaking operatives.
  • Motivation: Financially driven, specializing in data theft, extortion, and ransomware deployment.
  • Affiliations: Historically linked to ALPHV/BlackCat, recently affiliated with RansomHub and Qilin RaaS platforms.
  • Notable Attacks: MGM Resorts (2023, $100M loss), Caesars Entertainment (2023, $15M ransom paid), Okta, Coinbase, Reddit, and Snowflake customers.

TTPs for Ransomware Deployment:

  1. Reconnaissance:
    • Utilizes open-source intelligence (OSINT) platforms like LinkedIn to identify employees, roles, and organizational structures.
    • Scans for cloud infrastructure vulnerabilities, particularly in Microsoft Azure, AWS, and Google Workspace.
    • Registers typosquatted domains (e.g., victimname-sso[.]com) to mimic legitimate SSO portals.
  2. Initial Access:
    • Social Engineering: Employs SMS phishing (smishing), voice phishing (vishing), and MFA fatigue attacks to steal credentials. Impersonates IT helpdesk staff or new hires to trick employees into providing one-time passwords (OTPs) or resetting credentials.
    • SIM Swapping: Convinces cellular carriers to transfer a target’s phone number to a controlled SIM, bypassing MFA.
    • Phishing Campaigns: Deploys fake SSO login pages or leverages compromised credentials from public code repositories (e.g., GitHub).
  3. Persistence and Escalation:
    • Uses legitimate remote monitoring and management (RMM) tools (e.g., AnyDesk, TeamViewer, LogMeIn) to maintain access.
    • Exploits cloud environments, particularly Microsoft Azure, to escalate privileges by targeting identity administrators or misconfigured accounts.
    • Deploys scripts (e.g., PowerShell, Python) to steal credentials from tools like CyberArk or extract data from databases.
  4. Data Exfiltration and Extortion:
    • Exfiltrates sensitive data to cloud storage (e.g., MEGA[.]NZ) or U.S.-based data centers using tools like Rclone or Telegram.
    • Threatens to leak data unless a ransom is paid, often before deploying ransomware.
  5. Ransomware Deployment:
    • Historically used ALPHV/BlackCat ransomware, now affiliates with RansomHub and Qilin to encrypt systems, targeting VMware ESXi servers and Windows environments.
    • Communicates ransom demands via TOR, Tox, or encrypted apps, splitting proceeds with RaaS providers.

Strengths:

  • Sophisticated social engineering, leveraging native English proficiency and Western cultural familiarity.
  • Ability to evade detection by using living-off-the-land techniques and allowlisted applications.
  • Rapid adaptation of TTPs, shifting to new RaaS platforms post-ALPHV/BlackCat disruption.

Weaknesses:

  • Opportunistic rather than highly targeted, often abandoning environments after eviction.
  • Reliance on open-source tools and preformulated scripts, which are not highly novel.
  • Recent arrests (e.g., U.S. and U.K. members in 2024) may disrupt operations, though loose structure allows continuity.

2. DragonForce Overview

Profile:

  • Aliases: Limited open-source reporting; occasionally linked to pro-Palestinian hacktivist campaigns.
  • Origin: Suspected Middle Eastern or Eastern European base, with unclear leadership structure.
  • Motivation: Primarily financial, with potential ideological undertones targeting Western or Israeli-affiliated organizations.
  • Affiliations: Operates independently but may collaborate with initial access brokers on dark web forums.
  • Notable Attacks: Limited public attribution; linked to ransomware incidents in healthcare and manufacturing sectors (2023–2024).

TTPs for Ransomware Deployment:

  1. Reconnaissance:
    • Leverages dark web marketplaces to purchase stolen credentials or access logs.
    • Targets organizations with known vulnerabilities in VPNs, RDP, or unpatched software (e.g., CVE-2023-4966, Citrix Bleed).
  2. Initial Access:
    • Exploits compromised VPN credentials or unsecured RDP ports to gain entry.
    • Uses phishing emails with malicious attachments (e.g., DarkGate, PikaBot) to deliver payloads.
    • Purchases access from initial access brokers, focusing on high-value targets.
  3. Persistence and Escalation:
    • Deploys custom malware for command-and-control (C2), often using SOCKS proxies or VPN services like Tailscale.
    • Exploits privilege escalation vulnerabilities to gain administrative access.
    • Disables endpoint security tools by deleting Windows registry subkeys.
  4. Data Exfiltration and Extortion:
    • Exfiltrates data using tools like MEGAsync or Storage Explorer to external servers.
    • Employs double-extortion tactics, threatening to leak data alongside encryption.
  5. Ransomware Deployment:
    • Deploys proprietary ransomware variants, encrypting critical systems and demanding cryptocurrency payments.
    • Targets ESXi and Windows systems, similar to Scattered Spider, but with less reliance on RaaS models.

Strengths:

  • Stealthy operations with lower public visibility, complicating attribution.
  • Flexible targeting across industries, exploiting common vulnerabilities.
  • Potential ideological motivations may drive persistent attacks against specific geopolitical targets.

Weaknesses:

  • Limited open-source intelligence hinders detailed TTP analysis.
  • Smaller operational scale compared to Scattered Spider, potentially limiting resources.
  • Less sophisticated social engineering compared to Scattered Spider’s native English-speaking operatives.

3. Comparative Analysis

Aspect Scattered Spider DragonForce
Primary Motivation Financial (extortion, ransomware) Financial, possible ideological motives
Geographic Base U.S., U.K. (Western) Suspected Middle East/Eastern Europe
TTP Sophistication High (social engineering, cloud exploitation) Moderate (exploits, purchased access)
RaaS Affiliation Strong (ALPHV, RansomHub, Qilin) Limited; likely independent
Target Industries Tech, telecom, hospitality, finance, insurance Healthcare, manufacturing, geopolitically significant targets
Detection Evasion Advanced (living-off-the-land, allowlisted tools) Moderate (relies on stealth, less dynamic TTPs)
Law Enforcement Pressure High (FBI, U.K. arrests in 2024) Low (limited attribution)

Key Differences:

  • Scattered Spider’s strength lies in social engineering and cloud expertise, making it adept at targeting large, Western organizations with complex IT environments. DragonForce, while capable, relies more on exploiting known vulnerabilities and purchased access, suggesting a less resource-intensive approach.
  • Scattered Spider’s RaaS affiliations provide access to advanced ransomware strains, whereas DragonForce’s independent operations may limit its scalability but enhance its anonymity.

Potential Collaboration:

  • No evidence suggests direct collaboration between Scattered Spider and DragonForce. However, both groups operate in the same cybercriminal ecosystem, potentially sharing initial access brokers or dark web marketplaces. A partnership could combine Scattered Spider’s social engineering prowess with DragonForce’s stealth, creating a formidable threat.

4. Potential Ransomware Scenario

Target: A multinational financial services firm with cloud-based infrastructure (e.g., Microsoft Azure) and a large employee base.

Scattered Spider Attack Vector:

  1. Reconnaissance: Identifies IT helpdesk staff via LinkedIn and maps Azure configurations using OSINT tools.
  2. Initial Access: Sends smishing messages posing as the company’s SSO provider, tricking employees into entering credentials on a typosquatted domain (e.g., company-sso[.]com). Alternatively, uses vishing to impersonate IT staff, requesting OTPs.
  3. Persistence: Installs ScreenConnect or AnyDesk for remote access, escalating privileges via stolen CyberArk credentials.
  4. Exfiltration: Extracts customer data and financial records to MEGA[.]NZ using Rclone.
  5. Ransomware: Deploys RansomHub or Qilin ransomware, encrypting ESXi servers and demanding $5M in cryptocurrency. Communicates via Tox, threatening to leak data.

DragonForce Attack Vector:

  1. Reconnaissance: Purchases stolen VPN credentials from a dark web forum, identifying unpatched RDP ports.
  2. Initial Access: Exploits CVE-2023-4966 (Citrix Bleed) or uses compromised credentials to access the network.
  3. Persistence: Deploys a custom C2 beacon, disabling endpoint detection via registry modifications.
  4. Exfiltration: Uses MEGAsync to transfer sensitive data to an external server.
  5. Ransomware: Encrypts critical systems with a proprietary ransomware strain, demanding $2M. Threats include data leaks and potential public shaming if ideologically motivated.

Impact:

  • Operational Disruption: System downtime, affecting customer services and transactions.
  • Financial Loss: Ransom payments, recovery costs, and regulatory fines (e.g., GDPR, SEC).
  • Reputational Damage: Public exposure of data leaks, eroding customer trust.

5. Mitigation Strategies

  1. Employee Training:
    • Conduct regular phishing simulations and social engineering awareness training, emphasizing vishing and smishing red flags.
    • Train helpdesk staff to verify identities via secure channels, avoiding OTP sharing.
  2. Technical Controls:
    • Implement phishing-resistant MFA (e.g., FIDO2, hardware tokens) to counter SIM swapping and MFA fatigue.
    • Enforce strict RDP policies, closing unused ports and requiring VPN with MFA.
    • Monitor cloud configurations (e.g., Azure Policy, Security Center) for misconfigurations and unauthorized access.
    • Deploy endpoint detection and response (EDR) tools to detect lateral movement and anomalous activity.
  3. Network Segmentation:
    • Segment networks to limit lateral movement, isolating critical systems from user endpoints.
    • Use zero-trust architecture to enforce least privilege access.
  4. Incident Response:
    • Maintain offline backups and a recovery plan to restore systems without paying ransoms.
    • Monitor typosquatted domains and dark web forums for early warning of targeting.
  5. Collaboration:
    • Share threat intelligence with industry peers and authorities (e.g., FBI, CISA) to track Scattered Spider and DragonForce activities.
    • Engage with managed security providers for real-time threat hunting.

6. Conclusion

Scattered Spider and DragonForce represent evolving ransomware threats with distinct but overlapping capabilities. Scattered Spider’s social engineering expertise and RaaS affiliations make it a high-impact threat to large organizations, particularly in tech, finance, and hospitality. DragonForce, while less documented, poses a risk through stealthy exploitation and potential ideological motivations. Organizations must prioritize employee training, robust MFA, and cloud security to mitigate these threats. Ongoing law enforcement pressure on Scattered Spider may disrupt its operations, but its loose structure and DragonForce’s anonymity ensure continued ransomware risks. Proactive defense and intelligence sharing are critical to resilience.

Recommendation: Organizations should conduct immediate vulnerability assessments, enhance social engineering defenses, and align with CISA’s Cross-Sector Cybersecurity Performance Goals to reduce exposure.

Sources

– FBI and CISA Joint Cybersecurity Advisory (2023).
– EclecticIQ Ransomware in the Cloud Report (2024).
– Microsoft Threat Intelligence Report (2024).
– GuidePoint Security Scattered Spider Analysis (2024).
– Additional open-source intelligence from dark web monitoring and industry reports.

End of Assessment

next let’s see what they think about the trends of attacks like this:

Report on Trends in Attacks Blending Social Engineering and Ransomware

Report on Trends in Attacks Blending Social Engineering and Ransomware

Date: 28 April 2025

Prepared by: [Cybersecurity Analyst, Name Classified]

Classification: Restricted

Executive Summary

The convergence of social engineering and ransomware has emerged as a dominant trend in cybercriminal operations, exemplified by groups like Scattered Spider. These attacks combine sophisticated psychological manipulation with advanced ransomware deployment to infiltrate organizations, exfiltrate data, and extort substantial ransoms. Targeting industries such as technology, finance, hospitality, and critical infrastructure, these blended attacks exploit human vulnerabilities and technical weaknesses, resulting in significant financial, operational, and reputational damage. This report analyzes the trends, tactics, and impacts of such attacks, drawing on the methodologies of Scattered Spider, and provides recommendations to counter this evolving threat.

1. Overview of the Trend

Blended social engineering and ransomware attacks have surged in prevalence since 2022, driven by the following factors:

  • Sophistication of Social Engineering: Cybercriminals leverage native language proficiency, cultural familiarity, and open-source intelligence (OSINT) to craft highly convincing phishing, smishing (SMS phishing), and vishing (voice phishing) campaigns.
  • Ransomware-as-a-Service (RaaS) Proliferation: Platforms like ALPHV/BlackCat, RansomHub, and Qilin provide accessible, scalable ransomware tools, enabling groups to focus on initial access via social engineering.
  • Cloud and Remote Work Vulnerabilities: The shift to cloud-based infrastructures (e.g., Microsoft Azure, AWS) and remote work has expanded attack surfaces, with misconfigured accounts and remote access tools becoming prime targets.
  • Double-Extortion Tactics: Attackers combine data exfiltration with encryption, threatening to leak sensitive information unless ransoms are paid, increasing pressure on victims.

Key Statistic: According to the EclecticIQ Ransomware in the Cloud Report (2024), incidents involving social engineering as an initial access vector for ransomware increased by 40% from 2022 to 2024, with Scattered Spider linked to over 20% of high-profile cases.

2. Tactics, Techniques, and Procedures (TTPs)

Groups like Scattered Spider have refined a multi-stage approach to blend social engineering with ransomware deployment:

2.1 Reconnaissance

  • OSINT Utilization: Attackers use platforms like LinkedIn to identify employee roles, particularly IT and helpdesk staff, to tailor social engineering campaigns.
  • Infrastructure Scanning: Cloud environments (e.g., Azure, Google Workspace) are probed for misconfigurations or vulnerabilities.
  • Domain Spoofing: Typosquatted domains (e.g., victimname-sso[.]com) are registered to mimic legitimate single sign-on (SSO) portals, enhancing phishing credibility.

2.2 Initial Access via Social Engineering

  • Phishing Variants:
    • Smishing: SMS messages impersonate trusted entities (e.g., IT departments or SSO providers) to trick users into sharing credentials or one-time passwords (OTPs).
    • Vishing: Attackers pose as IT helpdesk staff or new hires, using native English proficiency to request OTPs or credential resets over the phone.
    • MFA Fatigue Attacks: Repeated MFA prompts overwhelm users, prompting them to approve unauthorized access.
  • SIM Swapping: Convincing cellular carriers to transfer a target’s phone number to a controlled SIM bypasses SMS-based MFA.
  • Fake SSO Portals: Compromised credentials are harvested via phishing pages mimicking corporate login interfaces, often hosted on typosquatted domains.

2.3 Persistence and Escalation

  • Legitimate Tools: Attackers deploy remote monitoring and management (RMM) tools like AnyDesk, TeamViewer, or ScreenConnect to maintain access, evading detection due to their allowlisted status.
  • Cloud Exploitation: Privilege escalation in cloud environments targets identity administrators or misconfigured accounts, often using stolen credentials from tools like CyberArk.
  • Script Deployment: PowerShell or Python scripts extract credentials or sensitive data from databases, facilitating lateral movement.

2.4 Data Exfiltration and Extortion

  • Exfiltration Tools: Sensitive data is transferred to cloud storage (e.g., MEGA[.]NZ) or U.S.-based data centers using tools like Rclone or Telegram.
  • Double Extortion: Attackers threaten to leak exfiltrated data unless a ransom is paid, often before deploying ransomware to maximize pressure.

2.5 Ransomware Deployment

  • RaaS Platforms: Groups leverage ransomware strains from ALPHV/BlackCat, RansomHub, or Qilin to encrypt systems, targeting VMware ESXi servers and Windows environments.
  • Communication Channels: Ransom demands are communicated via TOR, Tox, or encrypted apps, with proceeds split between the attacking group and RaaS providers.
  • High-Profile Examples: Scattered Spider’s 2023 attacks on MGM Resorts ($100M loss) and Caesars Entertainment ($15M ransom paid) demonstrate the scale of these operations.

3. Key Trends

3.1 Targeting High-Value Industries

  • Primary Sectors: Technology, telecom, hospitality, finance, and insurance are frequent targets due to their complex IT environments and valuable data.
  • Notable Incidents: Scattered Spider’s attacks on Okta, Coinbase, Reddit, and Snowflake customers highlight the focus on cloud-reliant organizations.

3.2 Evasion and Adaptation

  • Living-off-the-Land Techniques: Using legitimate tools (e.g., RMM software) and cloud services minimizes detection by traditional security solutions.
  • Rapid TTP Evolution: Following the disruption of ALPHV/BlackCat in 2024, groups like Scattered Spider swiftly pivoted to RansomHub and Qilin, demonstrating agility.
  • Law Enforcement Pressure: Despite arrests of Scattered Spider members in the U.S. and U.K. in 2024, the group’s loose structure enables continuity.

3.3 Financial and Operational Impact

  • Ransom Demands: Demands range from $2M to $15M, with additional costs from recovery, regulatory fines (e.g., GDPR, SEC), and downtime.
  • Reputational Damage: Data leaks erode customer trust, as seen in MGM Resorts’ 2023 breach.
  • Critical Infrastructure Risks: Targeting sectors like finance and telecom threatens front economic stability.

4. Case Study: Scattered Spider’s Attack Methodology

Scenario: Attack on a multinational financial services firm (2023).

Attack Vector:

  1. Reconnaissance: LinkedIn profiling identified IT helpdesk staff; Azure configurations were mapped using OSINT tools.
  2. Initial Access: Smishing campaign posed as the company’s SSO provider, directing employees to a typosquatted domain (e.g., company-sso[.]com). Vishing calls impersonated IT staff to obtain OTPs.
  3. Persistence: ScreenConnect was installed for remote access; stolen CyberArk credentials enabled privilege escalation.
  4. Exfiltration: Customer data and financial records were exfiltrated to MEGA[.]NZ using Rclone.
  5. Ransomware: RansomHub ransomware encrypted ESXi servers, with a $5M demand communicated via Tox.

Impact:

  • Operational: System downtime disrupted transactions for 48 hours.
  • Financial: $5M ransom paid, plus $10M in recovery costs.
  • Reputational: Public data leak led to a 15% stock price drop.

5. Mitigation Strategies

To counter the trend of blended social engineering and ransomware attacks, organizations should adopt a multi-layered defense:

  1. Employee Training:
    • Conduct regular phishing, smishing, and vishing simulations to raise awareness of social engineering tactics.
    • Train helpdesk staff to verify identities via secure channels, avoiding OTP sharing.
  2. Technical Controls:
    • Implement phishing-resistant MFA (e.g., FIDO2, hardware tokens) to mitigate SIM swapping and MFA fatigue.
    • Enforce strict RDP policies, closing unused ports and requiring VPN with MFA.
    • Monitor cloud configurations (e.g., Azure Policy, Security Center) for misconfigurations.
    • Deploy endpoint detection and response (EDR) tools to detect lateral movement.
  3. Network Segmentation:
    • Segment networks to limit lateral movement, isolating critical systems.
    • Adopt zero-trust architecture to enforce least privilege access.
  4. Incident Response:
    • Maintain offline backups and a recovery plan to avoid ransom payments.
    • Monitor typosquatted domains and dark web forums for early targeting indicators.
  5. Collaboration:
    • Share threat intelligence with industry peers and authorities (e.g., FBI, CISA).
    • Engage managed security providers for real-time threat hunting.

6. Conclusion

The blending of social engineering and ransomware, as exemplified by Scattered Spider, represents a highly effective and evolving cyberthreat. These attacks exploit human vulnerabilities and technical weaknesses, leveraging RaaS platforms and cloud environments to maximize impact. The trend’s growth, driven by sophisticated TTPs and high financial rewards, underscores the need for proactive defenses. Organizations must prioritize employee training, robust MFA, and cloud security while collaborating with industry and law enforcement to disrupt these operations. Immediate action, including vulnerability assessments and alignment with CISA’s Cybersecurity Performance Goals, is critical to resilience.

Sources

– FBI and CISA Joint Cybersecurity Advisory (2023).
– EclecticIQ Ransomware in the Cloud Report (2024).
– Microsoft Threat Intelligence Report (2024).
– GuidePoint Security Scattered Spider Analysis (2024).
– Open-source intelligence from dark web monitoring and industry reports.

End of Report

and now let’s get a table created with a historic view of key social engineering and ransomware attacks (this took 3 times to generate without unrelated incidents in, and even then this needs checking!)

Historical Cyber Attacks Blending Social Engineering and Ransomware

Historical Cyber Attacks Blending Social Engineering and Ransomware (2005–2025)

The following table lists significant cyber attacks over the last 20 years where social engineering was a confirmed initial access vector leading to ransomware deployment. Each entry details the attack, year, target, social engineering method, ransomware used, and impact. The Transport for London attack (2024) has been removed due to unconfirmed social engineering details, and WannaCry (2017) was previously excluded as it relied on the EternalBlue exploit, not social engineering.

Attack Year Target Social Engineering Method Ransomware Used Impact
CryptoLocker Attack 2013 Various organizations and individuals globally Phishing emails with malicious attachments posing as legitimate business documents CryptoLocker Infected thousands of systems, encrypted files with 2048-bit RSA key; attackers earned $27M in two months; average ransom $300–$600
Anthem Data Breach 2015 Anthem Inc. (U.S. healthcare provider) Spear phishing targeting employee credentials Ransomware (specific strain not publicly confirmed) Exposed 78.8M records; recovery costs estimated at $100M; significant reputational damage
Ubiquiti Networks Attack 2020 Ubiquiti Networks (U.S. tech company) Phishing email impersonating a trusted vendor Unknown ransomware strain Attackers demanded $2.5M ransom; data breach exposed customer information; $50M in recovery costs
Colonial Pipeline Attack 2021 Colonial Pipeline (U.S. critical infrastructure) Phishing email with compromised credentials DarkSide Paid $4.4M ransom; disrupted fuel supply across U.S. East Coast; $200M in economic losses
MGM Resorts Attack 2023 MGM Resorts (U.S. hospitality) Vishing impersonating IT helpdesk staff ALPHV/BlackCat (by Scattered Spider) $100M in lost bookings, $10M cleanup costs; systems down for days; customer data exfiltrated
Caesars Entertainment Attack 2023 Caesars Entertainment (U.S. hospitality) Social engineering via outsourced vendor ALPHV/BlackCat (by Scattered Spider) Paid $15M ransom; customer data leaked; ongoing dark web monitoring required
Indonesia National Data Center Attack 2024 Indonesia’s national data center Phishing targeting employee credentials LockBit Disrupted immigration services; data loss due to lack of backups; prompted nationwide data center audit

Notes

  • Selection Criteria: Attacks were chosen based on confirmed use of social engineering (e.g., phishing, vishing, smishing) as the initial access vector and ransomware deployment, with significant financial or operational impact.
  • Data Limitations: Exact ransomware strains may not always be publicly disclosed, particularly for older attacks like Anthem (2015).
  • Trends: Social engineering has grown more sophisticated, with phishing and vishing leveraging AI and deepfake technology in recent years. Ransomware-as-a-Service (RaaS) has lowered the barrier for such attacks since the 2010s.
  • Exclusions: WannaCry (2017) was excluded as it relied on the EternalBlue exploit, not social engineering. Transport for London (2024) was removed due to unconfirmed social engineering details, as public reports lack specific TTPs.

Sources

– EclecticIQ Ransomware in the Cloud Report (2024).
– Microsoft Threat Intelligence Report (2024).
– Arctic Wolf: History of Cybercrime (2024).
– Verizon Data Breach Investigations Report (2023).
– FBI Internet Crime Complaint Center (IC3) Reports (2021–2024).
– Chainalysis Ransomware Payment Reports (2024).
– CSIS Significant Cyber Incidents Timeline (2025).
– Additional open-source intelligence from industry reports.

End of Table

What other questions do you want it to answer? Do you think we are at a point where we can truly replace people with AI? I sure as hell do not!

Currently LLMs are tools, they look lovely, they are super fast but they are also sometimes inaccurate or miss major key data points.