Governance, Risk & Compliance · Singapore

Nikhil Menath Gopalkrishnan

Georgetown cybersecurity risk management graduate. I run risk assessments, control reviews and compliance tracking against ISO/IEC 27001, COBIT, NIST CSF and PCI DSS, with working knowledge of Singapore's IM8 and PDPA. I've handled confidential records in regulated healthcare and insurance settings, and I'm used to being the link between technical teams, operations and senior management.

ng791@georgetown.edu nikhilmg@gmail.com linkedin.com/in/nikhilgm Résumé

Experience

Student Worker, Office of Residential Life

2025 – 2026

Georgetown University, Washington, DC

  • Accessed and protected sensitive student information under university privacy and confidentiality rules.
  • Managed user access on the principle of least privilege, granting, updating and removing permissions as roles changed.

Software Engineering Team Lead

2023 – 2024

Bright Mind Enrichment and Schooling, New York

  • Assessed data privacy and security risks of LLM and generative AI integrations during pilot and rollout.
  • Acted as point of contact between programming teams and operations, coordinating issues, priorities and fixes.
  • Reported progress, risks and delivery status to senior management in clear, non-technical terms.
  • Built and hardened web applications on AWS using OWASP Top 10 secure coding practices.

COVID-19 Healthcare Assistant

2020 – 2021

Changi General Hospital, Singapore

  • Safeguarded confidential contact-tracing data in SafeEntry for the Ministry of Health under strict data protection controls.

Database Management

2019

Great Eastern Life, Singapore

  • Kept medical invoice records in Oracle/FPMS accurate and audit-ready at an MAS-regulated insurer.

Senior Trooper, 8th Battalion SIR

2016 – 2017

Singapore Armed Forces, Active Reserve

  • Enforced access control and physical security protocols at high-security installations alongside the Military Police Command.
  • National Service Excellence Award, Commander's Coin, Marksmanship Award.

Education

M.P.S., Cybersecurity Risk Management

2026

Georgetown University, Washington, DC

Information Assurance & Risk Management · Cybersecurity Governance Frameworks · Information Security Law & Compliance · Cyber Threat Intelligence

B.S., Information Technology

2023

New York Institute of Technology, New York

Network & Internet Security · Database Management · Cisco Networking Academy

Skills

Governance, risk & compliance

  • IT risk assessment
  • Compliance assessment
  • Control testing
  • Risk registers
  • Audit evidence
  • Remediation tracking
  • Governance reporting & metrics
  • Stakeholder coordination

Standards & frameworks

  • ISO/IEC 27001/27002
  • COBIT
  • PCI DSS
  • NIST CSF
  • NIST RMF / SP 800-53
  • SOC 2
  • SANS IR (PICERL)
  • OWASP Top 10
  • MITRE ATT&CK

Regulation & policy

  • Singapore IM8
  • Singapore PDPA
  • GDPR
  • CCPA

Data protection & security

  • Microsoft Purview
  • Data loss prevention
  • Access control
  • Incident escalation
  • Threat intelligence
  • Penetration testing
  • Secure coding

Tools

  • Jira
  • Excel
  • Python
  • SQL
  • AWS
  • Google Cloud
  • Oracle Database
  • Wireshark

Certifications

  • ISO/IEC 27001:2022 Information Security Associate (SkillFront)
  • Microsoft Copilot for Security
  • Google Cybersecurity Certificate
  • Google Cloud
  • CompTIA Security+ · in progress

Languages

  • English
  • Hindi
  • Tamil
  • Malayalam
  • French

Leadership

Vice President, Cybersecurity Society

2025 – 2026

Georgetown University

Organised cybersecurity networking events and hands-on security labs for the graduate community.

Cybersecurity Ambassador

2025

World Bank Group, Latin America Conference

Briefed international delegates on phishing and device security, and supported secure event operations and incident escalation.

Advisor of Operations, Presidential Cabinet

2025 – 2026

Georgetown Graduate Student Government

Projects

Ransomware response and recovery for an oil & gas operator

Jan – Apr 2026

Georgetown capstone · H. Ackme Oil & Gas case

A semester-long engagement on a ransomware attack that hit offshore SCADA and HMI systems during a federal investigation. I wrote the incident detection report on a firmware backdoor in third-party measurement devices, the vulnerability assessment, the incident response report and a change management policy under the Risk Management Framework, and briefed the board as CISO. With a partner, I co-authored the System Security Plan and the cyber surveillance plan.

  • NIST SP 800-53 Rev. 5
  • NIST SP 800-61
  • NIST SP 800-82
  • NIST SP 800-161
  • ISA/IEC 62443
  • RMF
Read the case studyClose

Cybersecurity Risk Management capstone · Georgetown University · January to April 2026 · Fictional company used for teaching

The situation

H. Ackme is an oil and gas operator running offshore drilling platforms. In the scenario, it found a backdoor in the firmware of measurement devices supplied by a third party, then suffered a ransomware attack that cut communications across several platforms and hit its SCADA and HMI systems. Operational and customer data were exposed. The FBI, DHS, CISA, TSA, FERC, BSEE and the Department of Defense opened a coordinated investigation, which put strict limits on how evidence had to be preserved during recovery. Ackme refused to pay the ransom and began a phased recovery.

What I delivered

  • Incident detection report to the CISO and board on the firmware backdoor, with a tiered monitoring and architecture plan.
  • Vulnerability assessment report covering technical, procedural and governance weaknesses, with prioritized fixes.
  • Incident response report walking through each phase from detection and containment to reporting, plus the ransomware mechanism, payload and regulatory consequences such as BSEE licensing.
  • Change management policy for the Change Control Approval Board, applying the Risk Management Framework to recovery changes.
  • Board briefing from the CISO's perspective, covering business impact, the threat actor, response actions and communication strategy.
  • System Security Plan and cyber surveillance plan, co-authored with a classmate.

Sample: vulnerability findings

An extract from the vulnerability assessment report, grouping what the attack exposed.

AreaFindingWhy it mattered
TechnicalPatch management gapsKnown vulnerabilities stayed open, increasing exposure to exploit-based attacks.
TechnicalInadequate network segmentationA flat network let the attackers move laterally and widened the blast radius.
TechnicalLimited endpoint visibilityWithout full EDR coverage, malicious activity was detected late.
TechnicalUntested backupsBackups existed, but weak testing left restoration time and data integrity uncertain.
ProceduralUnrehearsed response rolesIT, security, legal and leadership had never practised their roles together.
ProceduralUnclear reporting thresholdsIt was unclear when and how to notify regulators and law enforcement.
GovernanceThreat intelligence not used in risk decisionsIntelligence was not fed into enterprise risk management.
GovernanceWeak third-party oversightVendor access and controls were not reviewed continuously.

Sample: tiered detection roadmap

From the incident detection report, a plan that let the board choose investment against risk reduction.

TierRecommendationCostRisk reduction
1. FoundationsPassive OT network monitoring, OT logs centralized in the existing SIEM, firmware integrity checks.Low to moderateModerate cut in detection time
2. Detection and segmentationOT-specific intrusion detection, enforced segmentation and access control lists.ModerateMuch less lateral movement and operational impact
3. Zero trustZero trust across OT, with continuous authentication and device attestation.HighLarge drop in the likelihood and impact of a catastrophic incident

Recovery changes I put to the change board

  • Accelerated operating system and firmware upgrades.
  • Network segmentation and isolation of IT from OT.
  • Immutable, air-gapped backups.
  • Zero trust access controls with multi-factor authentication, plus privileged access management.
  • A complete asset inventory and configuration baseline.

Key takeaway

The central challenge was sequencing. Recovery changes had to be approved, documented and reversible while federal investigators still needed evidence preserved. A formal change process under the Risk Management Framework allowed fast recovery and a clean audit trail at the same time.

Enterprise risk management plan for a global device maker

Sep – Dec 2025

Information Assurance & Risk Management · Hoyas Inc. case

Redesigned the cyber risk management framework for a global device maker with hybrid cloud, OT/ICS networks and sensitive government contracts. The work covered framework selection, a crown-jewel assessment, control mapping to the five RMF steps, and a combined executive briefing for the CIO, CISO and CFO.

  • NIST CSF 2.0
  • ISO/IEC 27001
  • NIST RMF
Read the case studyClose

Information Assurance & Risk Management · Georgetown University · September to December 2025 · Fictional company used for teaching · Framework selection done in a team of four

The situation

Hoyas Inc. is a long-established maker of technical devices that now holds sensitive government contracts and depends on a global supply chain. Since its last risk assessment it had moved core systems to a hybrid cloud, connected manufacturing OT and ICS to the corporate network, and taken on a large remote workforce. Its risk framework no longer matched how the business ran.

Approach

I rebuilt the framework on NIST CSF 2.0, with ISO/IEC 27001 governance controls for accountability and auditability, so that executives, government customers and auditors could all trace how risk decisions were made. The process ran in five steps:

  • Identify the threats each asset is exposed to.
  • Identify the vulnerabilities those threats could exploit.
  • Evaluate the effect of exploitation on the business.
  • Rank assets by relative risk.
  • Define mitigations for the highest-risk assets first.

The crown jewels were government contract systems, identity and access management, manufacturing and OT systems, and core cloud platforms.

Key risks found

  • Lateral movement risk inside OT/ICS networks that lacked segmentation.
  • High-value executive and privileged accounts without strong controls, and limited MFA coverage.
  • Legacy systems with no compensating controls.
  • Cloud security baselines that were not consistently enforced.
  • No security baseline for mobile devices, plus third-party vendor risk.

Sample: control mapping

An extract from the control map I prepared for executives, showing how each control supports the framework and what it protects.

FunctionControlTypeWhat it protects
IdentifyVulnerability scanning and penetration testingDetective, technicalVisibility of weaknesses in network and application assets
ProtectMulti-factor authenticationPreventive, technicalAccounts against compromised passwords
ProtectNetwork segmentation (VLANs)Compensating, technicalCritical servers and databases, isolated from the general network
ProtectMandatory job rotationCompensating, operationalCritical functions against internal fraud and privilege abuse
DetectIDS and SIEMDetective, technicalNetwork traffic and logs, for real-time detection
DetectBaseline configuration auditsDetective, operationalWorkstations and servers against unauthorized changes
RespondIncident response planCorrective, operationalBusiness continuity, through clear containment steps
RecoverAutomated patch managementCorrective, technicalOperating systems and applications
RecoverFormal post-incident reviewCorrective, operationalLessons learned and training

The executive briefing

The plan ended with a combined briefing for the CIO, CISO and CFO, including a mitigation roadmap from day 0 to 180 and beyond. For the CFO, the business case in the scenario was an estimated investment of $1.8M in the fiscal year against $8–12M in avoided breach losses and an expected 45% reduction in incidents, while protecting the company's government contracts.

Vulnerability scanning and password-hash labs

Nov – Dec 2025

Hands-on labs · Tenable Nessus, Hashcat, John the Ripper

Ran a Nessus Essentials scan of a macOS host, then triaged five findings, from a critical Chrome issue to Ruby Rack CVE-2025-61770, with impact and remediation for each. In a separate lab I cracked MD5, SHA-1, SHA-256, sha512crypt and bcrypt hashes and compared how well each algorithm resists attack.

  • Nessus
  • Hashcat
  • John the Ripper
  • Python

Supply chain risk action memo for the Federal Acquisition Security Council

Jul – Aug 2025

Information Security Laws & Compliance · ACME Company case

Writing as the council's executive director, I assessed foreign-influence and supply chain risk from a vendor with an overseas parent, and recommended that agencies move off the vendor under FASCSA orders. Our team also prepared the council's initial meeting briefing, and I added a general counsel view on sharing incident information with the federal government.

  • FASCSA
  • FAR
  • Supply chain risk

Cybersecurity governance programme for a manufacturer

Jun – Aug 2025

Cybersecurity Governance Frameworks · Parts Unlimited case

Built a governance programme over a summer. It includes a framework recommendation memo for the CEO, a workforce governance comparison of DoD and NIST models, an incident response policy with a RACI matrix, and a final plan that adopts NIST CSF through applied strategic planning and Kotter's change model.

  • NIST CSF
  • NIST SP 800-61
  • RACI
  • Change management

National cybersecurity strategy for Laos

Jun 2025

Team consulting presentation · Southeast Asia policy

A phased reform roadmap for Laos that uses ASEAN support mechanisms, regional legal alignment and Thailand as a model, covering foundation building through to legal and policy development.

  • ASEAN
  • GDPR
  • National policy

Writing

The convergence of biometrics and blockchain

Dec 2025

Why passwords and centralised identity stores keep failing, and whether pairing biometric authentication with decentralised records can make identity verification more private and harder to steal.

Read the full pieceClose

Cyber Threat Intelligence · Georgetown University · 9 December 2025

Overview

Emerging technologies are often disruptive by nature. As society digitises at an accelerating pace, it needs identification systems that are secure, private and trustworthy. Conventional approaches such as passwords, centralized identity databases and tokens keep failing in the face of large-scale breaches, credential theft and identity fraud. As online ecosystems grow, identity verification needs to grow at the same pace, or ideally faster.

Biometrics and blockchain are usually discussed separately, and both play important roles in cybersecurity. They are now converging into a new approach that could reshape how identity is managed and verified. The idea is still at an early stage, and it brings real promise along with significant challenges that deserve critical evaluation.

Biometrics

Biometrics identify people by physical traits such as fingerprints, faces, irises and gait, or by behavioral patterns. They offer clear advantages: they support non-repudiation, they are convenient, and they reduce reliance on passwords. The same properties that make biometrics powerful also make them a liability. A stolen fingerprint cannot be reset, and facial templates stored on centralized servers become a prime target for threat actors. Surveillance and misuse are serious concerns too, because people rarely know who holds their biometric data or how it is used. The core question is how to keep the convenience and strength of biometrics while removing the risks of storing them centrally.

Blockchain

Blockchain is decentralized and tamper-evident, and it is built on cryptographic proofs. It offers a model of trust that does not depend on a central authority, and its immutability and distributed consensus make it well suited to auditability and integrity. Applied to identity, blockchain enables self-sovereign identity (SSI), where individuals control their own identity attributes instead of relying on a central identity provider. Blockchain was never designed to store large amounts of sensitive data, though. The immutability that makes it secure also makes it dangerous for personal information, which may need to be corrected or deleted.

Where the two converge

Combined, biometrics and blockchain point toward identity systems that are decentralized, privacy-preserving and user-controlled, as well as tamper-resistant, auditable and firmly bound to a real person. That makes the convergence especially promising for healthcare, supply chains, defense, finance and cross-border travel. All of these are high-risk domains that need strong identity assurance.

An emerging architecture

Most proposed designs avoid storing raw biometric data on-chain and are privacy-preserving by design. They typically work in five steps:

  • Capture. A device such as a smartphone captures the biometric and converts it into a template.
  • Local encryption. The template is encrypted on the device, and only a cryptographic hash or commitment is posted to the blockchain.
  • Verification anchor. At authentication, a fresh scan is compared off-chain and checked against the hash stored on-chain.
  • Zero-knowledge proofs. ZKPs prove that a biometric match took place without revealing the biometric itself.
  • Self-sovereign identity wallets. Users hold their biometric-bound credentials on their own device, while the blockchain provides a public, tamper-proof trust layer.

The model combines the integrity of blockchain with the strong identity binding of biometrics, without exposing sensitive data.

Emerging uses

The pairing shows promise across digital identity, physical and IoT access, financial services, and healthcare access and consent.

  • Digital identity. Many countries are exploring decentralized digital identity. The EU's eIDAS 2.0, for example, is examining how biometrically anchored credentials could improve cross-border authentication and reduce the risk of data leaks.
  • Healthcare. Patients could authenticate with biometrics bound to blockchain-based consent records, making the sharing of medical data auditable, user-controlled and tamper-proof.
  • Financial services. Banks and fintech platforms could cut down on repeated Know Your Customer (KYC) checks, reducing fraud and streamlining onboarding.
  • Access control. Biometric blockchain identity could secure access to facilities, connected devices and even vehicles, removing the central points of failure that attackers often exploit in IoT ecosystems.

Conclusion

Integrating biometrics with blockchain has long interested me. Open questions remain, including how these systems will hold up against quantum computing, but the area keeps proving worth exploring. Implementation will take substantial research to make sure the technology does more good than harm. I plan to keep studying the topic over the coming months, with the aim of finding better answers and applying them in future projects.

References

  • Antonopoulos, A. M., & Wood, G. (2018). Mastering Ethereum: Building smart contracts and DApps. O'Reilly Media.
  • Jain, A. K., Ross, A. A., & Nandakumar, K. (2011). Introduction to biometrics. Springer. https://doi.org/10.1007/978-1-4614-0797-9
  • Zyskind, G., Nathan, O., & Pentland, A. (2015). Decentralizing privacy: Using blockchain to protect personal data. 2015 IEEE Security and Privacy Workshops, 180–184. https://doi.org/10.1109/SPW.2015.27

APT41: China's dual espionage and cybercrime actor

Nov 2025

An assessment of APT41's history, motivations and TTPs, with a recent campaign tied to Chinese national and economic goals. It builds on a team briefing for incoming political appointees, where I covered the Department of Defense perspective.

Read the full pieceClose

Cyber Threat Intelligence · Georgetown University · 3 November 2025

Abstract

This paper assesses APT41, also tracked as Double Dragon, Barium and Winnti, a prolific China-linked threat actor active since at least 2012. It summarizes the group's history, motives and tactics, techniques and procedures (TTPs), assesses its targeting and sophistication, and uses a recent campaign as a case study connecting APT41's tradecraft to China's national, economic and strategic objectives.

History and strategic motivation

APT41 is known for a dual operational profile: state-aligned espionage alongside financially motivated intrusions, both attributed to the same cluster of actors and tools. Security firms and government reports assess with high confidence that parts of APT41 have collected intelligence in line with Chinese state priorities, including technology theft and espionage against healthcare and telecommunications. Other activity has targeted personal or criminal profit, particularly in the video game industry. This "moonlighting" complicates attribution. It also reflects a persistent pattern in PRC operations, in which deniable contractors and non-state actors are used to advance economic and technology goals.

TTP evolution and sophistication

APT41 has evolved from opportunistic phishing and commodity malware to highly targeted operations that use:

  • Supply chain compromises. Injecting malicious code into legitimate software updates gives the group broad, stealthy access across many target environments.
  • Stealthy persistence. Sophisticated backdoors, stolen digital certificates and bootkits help it stay in place and evade detection.
  • Rapid exploitation. Exploiting high-impact public vulnerabilities such as Log4j gives it initial access to a wide range of victims.

APT41 blends its tradecraft, reusing espionage-grade malware in both state and criminal operations. Its capabilities point to medium-to-high technical sophistication. The group chooses victims carefully, tolerates long dwell times, and uses supply chain access to extend its reach while lowering the risk of attribution.

Targeting and national objectives

APT41's focus on healthcare, high-tech manufacturing, government and higher education matches the PRC's strategic targets. Those targets serve several aims: acquiring dual-use intellectual property, gathering negotiating and political intelligence, and supporting industrial policy such as "Made in China 2025." By targeting research institutions and stealing research and development data from companies, APT41 supports economic modernization and military-civil fusion at the same time. Its habit of exploiting niche third-party software shows an emphasis on weak links as a way into high-value networks. This is consistent with an intelligence-driven, resource-efficient approach to national objectives.

Case study: the USAHERDS campaign (2021–22)

Mandiant documented APT41 exploiting vulnerabilities in USAHERDS, a livestock management application, and in other ASP.NET web applications to compromise several U.S. state government networks. The actor abused hard-coded keys and repeatedly exploited application flaws to move from a single vendor application into state infrastructure, gaining persistent access and the ability to move laterally. The campaign shows both APT41's operational logic and its strategic ends:

  • Leverage through shared software. By targeting sector-specific software used by many government bodies, one third-party weakness gave the group access to many victims at once.
  • Strategic intelligence value. State government networks hold large volumes of personal data along with agricultural, trade and supply chain information, all relevant to trade negotiations and policy decisions.
  • Opportunism. APT41 pairs technical creativity with opportunism, weaponizing both publicly known and novel vulnerabilities to keep its access.

The nature of its tradecraft

APT41's sophistication shows in three areas:

  • Patience and tailoring. Tailored spear phishing, persistence mechanisms and multistage exploitation point to mature tradecraft and strong operational security.
  • Supply chain operations. Code injection, certificate abuse and developer compromise require deep knowledge and set APT41 apart from commodity cybercrime.
  • Hybrid operations. Reusing tools across state and profit-driven campaigns suggests access to privileged tooling and a blurred line between sanctioned state activity and crime.

APT41 seeks intellectual property and intelligence for negotiations and policy, all in service of national technological advancement. Its shift from phishing and commodity malware to certificate-based persistence reflects the growing payoff from targeting shared services and software providers rather than individual end-user organizations.

Conclusion

APT41 is a model of the modern PRC-linked threat actor. It runs both state and criminal operations, and its objectives are clear: to advance China's economic and intelligence goals. Its move toward exploiting supply chains and applications has made it more efficient and extended its reach. Third-party and developer compromises let it gain broad access while still running precise espionage missions. To defend against APT41 and similar actors, organizations should prioritize supply chain security and sound vendor management, and patch internet-facing applications quickly to remove the advantages these actors exploit.

References

  • FireEye. (2019). APT41: A dual espionage and cyber crime operation. https://services.google.com/fh/files/misc/apt41-a-dual-espionage-and-cyber-crime-operation.pdf
  • Google Cloud Threat Intelligence. (2024, July 18). APT41 has arisen from the DUST. https://cloud.google.com/blog/topics/threat-intelligence/apt41-arisen-from-dust
  • MITRE ATT&CK. (n.d.). APT41 (G0096). https://attack.mitre.org/groups/G0096/
  • U.S. Department of Health and Human Services, HC3. (2023). China-based threat actors: Threat profile. https://www.hhs.gov/sites/default/files/china-based-threat-actor-profiles-tlpclear.pdf
  • Wired. (2022). Chinese spies hacked a livestock app to breach U.S. state networks. https://www.wired.com/story/china-apt41-hacking-usaherds-log4j

APT29: profiling a Russian state espionage group

Oct 2025

Who APT29 targets, how its operations serve Russian foreign-intelligence requirements, and what defenders in government and diplomatic sectors should watch for.

Read the full pieceClose

Cyber Threat Intelligence · Georgetown University · 1 October 2025

Background

APT29 is a state-sponsored Russian espionage group attributed to Russia's Foreign Intelligence Service (SVR). Since 2008 it has targeted Western governments, defense organizations, diplomatic missions, NGOs, think tanks and IT service providers, in line with Russia's foreign policy and intelligence requirements. Its operations aim to give Russia military and diplomatic situational awareness and to inform its responses on energy policy and sanctions. Taken together, its work is about strengthening Russia's bargaining position, especially in its prolonged confrontation with the West over Ukraine.

Motivations and objectives

APT29 is not mainly interested in disruption. It is focused on collecting intelligence, and its core intent is persistent, covert access to high-value networks and cloud identities so it can extract sensitive policy, security and technology information. It has three main objectives:

  • Gaining pre-decisional insight into U.S. and European policy.
  • Reaching partners' communications and credentials across cloud and federated environments.
  • Learning defenders' telemetry and investigative methods, so it can harden future operations.

Government and industry attributions consistently describe APT29 as a persistent threat group with an espionage mission.

Core TTPs and how they have changed

Historically, APT29 has excelled at spear phishing, stealthy "living off the land" persistence and strategic web compromise, often using legitimate services so its traffic blends in with normal activity. Since 2020 it has shifted noticeably toward identity abuse and cloud-based initial access. Its methods now include:

  • Password spraying against Entra ID (formerly Azure AD).
  • Token theft through adversary-in-the-middle phishing.
  • Manipulating OAuth applications and consent grants.
  • Exploiting trusted relationships to reach downstream tenants.

Targeting

With its intelligence focus, APT29 prioritizes foreign ministries and policymakers, the defense industrial base, NGOs, multilateral institutions, academics researching Russia, and technology firms that support digital infrastructure. These priorities map directly onto Russian state interests: how sanctions are designed and enforced, weapons transfers, energy markets, and security aid to Russia's adversaries, such as Ukraine.

Case study: Midnight Blizzard and Microsoft

On 12 January 2024, Microsoft detected a nation-state intrusion by APT29, which it tracks as Midnight Blizzard. The actor had used password spraying against a legacy, non-production test tenant that lacked multi-factor authentication (MFA). It compromised a small number of corporate email accounts, including members of senior leadership and the security and legal teams, and exfiltrated emails and attachments. In March 2024, Microsoft reported that the same actor was using information from the stolen emails to seek further access, including to source code repositories and internal systems, while increasing the volume of its password spraying. Microsoft said it had found no evidence that customer-facing systems were compromised.

Why it matters strategically

A breach at a company like Microsoft causes compromise on many levels. Identity, cloud infrastructure and email form the backbone of governments and enterprises, so access to them serves Russia's intelligence needs directly. Reading executive correspondence and threat intelligence workflows gives insight into how the West defends against Russian operations. It can also expose credentials and secrets shared with customers, and offers a vantage point for mapping trust and authentication patterns between organizations. All of this fits the SVR's goal of expanding collection quietly while limiting its adversaries' visibility into its tradecraft. The campaign reinforced the SVR's broader move toward identity attacks for initial cloud access, instead of noisy endpoint exploitation.

Tactics observed

  • Initial access. Large-scale password spraying against cloud accounts, plus targeted spear phishing.
  • Credential and token abuse. Malicious OAuth application registrations and phishing to create durable backdoors in cloud tenants.
  • Lateral movement and persistence. "Living off the cloud" by abusing legitimate APIs, trusted relationships and the supply chain.
  • Collection and exfiltration. Executive mailboxes, internal communications and code repositories, with stolen data used to enable later intrusions.

Assessment of sophistication

APT29 shows a high level of sophistication in its identity-focused operations: disciplined operational security, use of built-in cloud features to avoid detection, and patient, multistage intrusion chains. The Microsoft incident also shows how the group gets high returns from low-cost techniques, in this case a password spray that worked because MFA was missing. Its efficient, stealthy approach is consistent with SVR tradecraft.

Compared with its earlier spear phishing and stealthy backdoors, APT29 has shifted sharply toward password spraying and weaknesses in cloud environments, particularly cloud identities. A February 2024 joint advisory from the UK's NCSC and CISA urged defenders to harden cloud identity paths, deploy MFA effectively, monitor OAuth applications and restrict consent grants so information stays on a need-to-know basis. Microsoft's later disclosures also describe continued spear phishing.

Risk outlook

APT29 will likely keep collecting intelligence on Western policy, technology providers and defense organizations, with even more emphasis on cloud identity as a route in. Russia's strategic isolation and wartime intelligence needs suggest its espionage tempo will stay high. Organizations with geopolitical relevance should assume they are targets.

Recommendations for defenders

  • Require phishing-resistant MFA.
  • Retire legacy tenants and legacy authentication protocols.
  • Apply role-based access control and least privilege to OAuth applications.
  • Monitor continuously, especially identity and consent activity.

References

  • Cybersecurity and Infrastructure Security Agency & National Cyber Security Centre. (2024, February 26). SVR cyber actors adapt tactics for initial cloud access (AA24-057A).
  • Microsoft Security Response Center. (2024, January 19). Microsoft actions following attack by nation state actor Midnight Blizzard.
  • Microsoft Security Response Center. (2024, March 8). Update on Microsoft actions following attack by nation state actor Midnight Blizzard.
  • Microsoft Threat Intelligence. (2024, January 25). Midnight Blizzard: Guidance for responders on nation-state attack.
  • Microsoft Threat Intelligence. (2024, October 29). Midnight Blizzard conducts large-scale spear-phishing campaign using RDP files.
  • MITRE ATT&CK. (n.d.). APT29 (G0016). https://attack.mitre.org/groups/G0016/

Change Healthcare ransomware: an executive memo

Sep 2025

A briefing for leadership on the ALPHV/BlackCat breach of a UnitedHealth subsidiary: how the attackers got in through a remote access portal, what it disrupted across US healthcare, and what to do about it.

Read the full pieceClose

Communication Strategy for Information Security Professionals · Georgetown University · 22 September 2025 · Written in role as a cybersecurity operations manager

To: Executive Leadership Team
Subject: Executive summary and recommendations on the Change Healthcare ransomware breach

Background

Change Healthcare, a subsidiary of UnitedHealth Group, suffered a ransomware attack that disrupted a large part of U.S. healthcare infrastructure. The attackers were affiliated with the ALPHV (BlackCat) ransomware group. They used stolen credentials to log in to a Citrix remote access portal that did not have multi-factor authentication (MFA). Over the following nine days they moved through the network before deploying ransomware that encrypted systems supporting pharmacy claims, payment processing and eligibility checks.

UnitedHealth estimated the cost of the disruption at more than $2.3 billion, and later investigation found that the personal information of about 190 million people had been exposed. Beyond the financial and reputational damage, the outage disrupted patients' access to care. Federal agencies had publicly warned beforehand that ransomware operators were targeting healthcare systems and that attacks were becoming more frequent.

Issue 1: Weak authentication and remote access controls

The attack was made possible mainly by compromised credentials and the lack of MFA on a critical remote access system. That gap let the attackers walk past weak defenses.

Recommended actions:

  • Deploy phishing-resistant MFA on all remote access paths.
  • Implement role-based and attribute-based access controls.
  • Add a remote-access-compromise playbook to the incident response plan, covering forced credential resets and revocation of active sessions.

Issue 2: Lateral movement and inadequate segmentation

Once inside, the attackers moved quickly across the network to core systems. This exposed insufficient network segmentation and weak monitoring of lateral movement.

Recommended actions:

  • Strengthen segmentation, starting with critical systems.
  • Deploy endpoint detection and response (EDR) across the organization to detect anomalies quickly.
  • Upgrade recovery capability by maintaining offline backups.
  • Develop an enterprise ransomware playbook with clear procedures for restoring critical systems.

Improving incident response

To build resilience, I recommend that we:

  • Run regular tabletop exercises that simulate identity compromise and ransomware scenarios.
  • Update vendor contracts to require MFA, real-time breach notification and log sharing.
  • Staff a team to monitor critical systems around the clock.
  • Keep prepared communication templates for customers, regulators and the media to limit reputational damage.
  • Track incident response metrics continuously.

Conclusion

The Change Healthcare breach shows how a single weakness, such as one remote portal without MFA, can escalate into an industry-wide crisis costing billions of dollars. By closing gaps in authentication and network segmentation and strengthening our incident response plans, we can reduce both the likelihood and the impact of a similar event.

References

  • CISA, FBI, & HHS. (2024). #StopRansomware: ALPHV Blackcat (AA23-353A). Cybersecurity and Infrastructure Security Agency. https://www.cisa.gov/
  • Kapko, M. (2024, April 30). Change Healthcare compromised by stolen credentials, no MFA. Cybersecurity Dive. https://www.cybersecuritydive.com
  • Reuters. (2025, January 24). UnitedHealth says hack impacted ~190 million people. https://www.reuters.com

Connected car cybersecurity

Aug 2025

How ISO/SAE 21434 and ISO 24089 apply to vehicles whose safety now depends on code running across ECUs, mobile apps, cloud APIs and dealer portals.

Read the full pieceClose

Information Security · Georgetown University · 17 August 2025

Abstract

Vehicles are increasingly defined by software. Their safety and security depend on code that runs across electronic control units (ECUs), mobile apps, cloud APIs and dealer portals. Standards such as ISO/SAE 21434 (cybersecurity engineering) and ISO 24089 (software update engineering) work alongside guidance such as NHTSA's best practices, which call for over-the-air (OTA) update readiness, proper logging and network segmentation. Even so, high-profile incidents keep occurring. The 2015 Jeep Cherokee hack and the web and API weaknesses researchers found at carmakers in 2023 and 2024 show that best practices do not guarantee fast remediation across vehicles already on the road.

I propose that regulators require a "Digital Recall" mechanism. It would combine secure-by-design architecture with real-time monitoring, the ability to quarantine connections, industry-wide Software Bill of Materials (SBOM) transparency, and OTA hot patching that does not require a reboot. These measures are complex to build, but they are proactive and could keep a vehicle safe to drive when threats are detected. Treating cyber defects like mechanical recalls, with firm timetables and public reporting, would significantly reduce systemic risk.

Why change is necessary

Modern vehicles integrate many ECUs connected over CAN, FlexRay or automotive Ethernet and linked to telematics units and cloud services. As carmakers move to centralized and zonal computing, the code base and the attack surface keep growing. Embedded security has improved significantly, but the weak links now sit off-board, in web portals, mobile app backends and dealer tools. Poor authorization, unscoped tokens and missing rate limits in carmakers' APIs make remote attacks on vehicles more likely. A compromised update pipeline could push malicious firmware to entire fleets, putting safety, customer trust and the business at risk.

Standards help, but they are process- and audit-oriented and applied unevenly. They do not enforce time-to-mitigation, public transparency, defect detail or coverage by Vehicle Identification Number (VIN). A Digital Recall could fill those gaps.

Lessons learned

  • Jeep Cherokee, 2015. Researchers remotely reached safety-critical functions such as steering and braking through the Uconnect infotainment system. The case led to the first cybersecurity recall, of about 1.4 million vehicles, and focused the industry on segmentation and secure gateways.
  • Carmaker web and API flaws, 2023–24. Researchers showed that vehicles could be affected without compromising any ECU. In 2024, Kia vehicles could be located and controlled remotely using only a licence plate, through weaknesses in cloud-side systems.

These incidents show that security has to cover the whole system of systems: in-vehicle networks, mobile and web apps, OTA pipelines, cloud services and dealer tools.

Outlook for 2025 to 2030

The industry's direction and recent findings suggest risk is shifting toward cloud and API systems as telematics, mobile keys, remote diagnostics and in-car commerce expand. The shift is happening less because embedded security has improved and more because off-board services are growing faster, and a flaw there can affect whole fleets at once. That points to more investment in API security engineering, identity and authorization logging, and anomaly detection beyond the vehicle, along with stronger guarantees for OTA updates and code signing.

A blueprint for Digital Recall

The goal is to make remediating cyber defects mandatory and measurable.

ElementWhat it involves
Secure by designIsolate safety-critical domains behind secure gateways, deny traffic between infotainment and drive domains by default, and keep secure logs with IDS/IPS support for forensics. Follow international standards for threat analysis and risk assessment, keep supplier evidence trails, and run system-wide red-team exercises, not just ECU testing.
SBOM transparencyShare SBOMs securely with regulators so carmakers and regulators can quickly assess exposure to a new vulnerability and identify affected vehicles by VIN.
OTA hot patchingUse Uptane-style multi-role signing with rollback protection and staged rollouts for faster, safer fixes, with regular audits against service-level expectations.
Safe degradeWhen IDS or cloud analytics detect a compromise, quarantine external connections, tell the owner the vehicle needs urgent digital service, and accept only emergency signed updates until integrity is restored.
Post-quantum roadmapPlan certificate management, message security and misbehavior detection so the fleet is ready as post-quantum cryptography arrives.
Implementation0–12 months: implement the blueprint, audit regularly and keep regulators informed. 12–24 months: certify the cybersecurity and software update management systems (CSMS and SUMS) required under UN R155 and R156, set internal time-to-mitigation targets and pilot safe-degrade-by-default. 24–36 months: commission annual external audits of the Digital Recall function.

Challenges

Any new approach will meet objections, and change has to be managed so that the workforce stays on board. One common objection is that SBOMs help attackers. SBOMs list components, not algorithms, and they speed up the response to new CVEs. NTIA and CISA have published guidance on handling confidentiality concerns. Another objection concerns the risk of hot patches and OTA updates. Leaving vehicles unpatched is riskier than using Uptane-style frameworks with staged rollouts and rollback protection.

Conclusion

Connected cars are now part of public safety infrastructure, so problems need to be contained quickly and verifiably. Using SBOMs, Uptane-class OTA updates, safe-degrade-by-default design and recall-style public reporting, alongside standards such as ISO/SAE 21434 and ISO 24089, regulators and carmakers can deliver robust, reliable cybersecurity. As risk moves toward the cloud and API perimeter, Digital Recalls would help carmakers close the gap between discovering a flaw and protecting drivers and the public.

References

  • Auto-ISAC. (2025, April 2). Auto-ISAC releases updated "Best Practice Guides" for the automotive community. https://automotiveisac.com/press-news/
  • Avertium. (2023, February 16). Threat actors, car systems, and weak API security. https://www.avertium.com/resources/threat-reports/tir-20230213-threat-actors-car-systems-and-weak-api-security
  • CISA. (2024, October 15). Establishing a common software bill of materials (SBOM). https://www.cisa.gov/
  • Curry, S. (2023, January 3). Web hackers vs. the auto industry. https://samcurry.net/web-hackers-vs-the-auto-industry
  • Curry, S. (2024, September 20). Hacking Kia: Remotely controlling cars with just a license plate. https://samcurry.net/hacking-kia
  • Greenberg, A. (2015, July 21). Hackers remotely kill a Jeep on the highway, with me in it. Wired. https://www.wired.com/2015/07/hackers-remotely-kill-jeep-highway/
  • IEEE. (2016). IEEE 1609.2: Security services for applications and management messages. https://standards.ieee.org/ieee/1609.2/6038/
  • ISO. (2023). ISO 24089: Road vehicles, software update engineering. https://www.iso.org/standard/77796.html
  • ISO/SAE. (2021). ISO/SAE 21434: Road vehicles, cybersecurity engineering. https://www.iso.org/standard/70918.html
  • NHTSA. (2022, September 7). Cybersecurity best practices for the safety of modern vehicles. https://www.nhtsa.gov/
  • NTIA. (2021, July 12). The minimum elements for a software bill of materials (SBOM). https://www.ntia.gov/report/2021/minimum-elements-software-bill-materials-sbom
  • Twardokus, G., et al. (2024). Practical post-quantum authentication for V2V. NDSS Symposium. https://www.ndss-symposium.org/
  • UNECE. (2021). UN Regulation No. 155 (cybersecurity and CSMS) and No. 156 (software update and SUMS). https://unece.org/
  • Uptane. (n.d.). A secure software update framework for vehicles. https://uptane.org/

GDPR and the US sectoral approach to privacy

Jul 2025

A comparison of the EU's single comprehensive regulation with the US patchwork of sector laws, and what each means for organisations handling cross-border data.

Read the full pieceClose

Information Security · Georgetown University · 16 July 2025

Introduction

In the digital age, protecting personal data has become a cornerstone of cybersecurity and public trust. National policies and global data flows shape privacy protection in very different ways around the world, and the contrast is clearest between the European Union's General Data Protection Regulation (GDPR) and the United States' sectoral approach. The GDPR is a single, comprehensive legal framework, while the U.S. relies on a fragmented mix of federal and state laws aimed at particular sectors and types of data. This paper analyzes both models, evaluates how effective they are and how they are implemented, and considers what they mean for individuals and organizations. Comparing their strengths and weaknesses also raises the question of whether global harmonization of privacy law is desirable or feasible.

The General Data Protection Regulation

The GDPR was adopted in 2016 and became enforceable in May 2018 to protect the personal data of people in the European Union. Unlike sectoral models, it applies to any organization that processes the personal data of people in the EU, whatever its sector or size, and including organizations based outside the EU (European Parliament, 2016).

Article 5 sets out the GDPR's core principles: lawfulness, fairness and transparency; purpose limitation; data minimization; accuracy; storage limitation; integrity and confidentiality; and accountability. The regulation gives data subjects extensive rights to access, correct, erase and port their data. It also requires organizations to notify the supervisory authority of a personal data breach within 72 hours of becoming aware of it, and to inform affected individuals without undue delay when the breach poses a high risk to them.

National data protection authorities (DPAs), coordinated by the European Data Protection Board (EDPB), oversee compliance. Organizations that break the rules face fines of up to €20 million or 4% of annual global turnover, whichever is higher. Because of its extraterritorial reach, the GDPR is widely treated as the global benchmark for data protection, and countries including India (DPDPA) and Brazil (LGPD) have adopted laws with similar approaches.

The U.S. sectoral approach

The United States has no single, comprehensive federal privacy law. It relies on a patchwork of laws that vary by sector and by state. The Federal Trade Commission (FTC) is the main regulator on privacy matters, enforcing Section 5 of the FTC Act, which prohibits "unfair or deceptive" practices (Solove & Hartzog, 2014).

Key federal laws:

  • HIPAA (Health Insurance Portability and Accountability Act) protects health information.
  • COPPA (Children's Online Privacy Protection Act) protects the data of children under 13.
  • FERPA (Family Educational Rights and Privacy Act) governs student education records.
  • GLBA (Gramm-Leach-Bliley Act) covers consumer financial information.

State laws, for example:

  • CCPA (California Consumer Privacy Act) gives California residents rights over their personal information.
  • CPRA (California Privacy Rights Act) amends and expands the CCPA.

Together these laws form a fragmented system, with protections, enforcement methods and compliance demands that differ from one law to the next. The complexity confuses consumers and places a heavy compliance burden on organizations operating across several jurisdictions.

Depth and breadth of implementation

The GDPR is comprehensive and uniform across Europe. Each member state enforces it through its national DPA, with the EDPB supporting consistent interpretation. Beyond data protection, the GDPR has shaped the EU's wider digital strategy, including the Digital Services Act and the AI Act. Implementation in the United States, by contrast, is uneven. HIPAA and GLBA are well established in their domains, but they leave significant gaps outside healthcare and finance. To fill those gaps, states such as California, Colorado, Connecticut, Utah and Virginia have passed comprehensive privacy laws, but federal harmonization remains uncertain.

Contribution to control and threat reduction

The GDPR aims to reduce vulnerabilities through principles such as purpose limitation, data minimization and privacy by design, which push organizations to consider privacy risk proactively. Data Protection Impact Assessments (DPIAs) and the appointment of Data Protection Officers (DPOs) build risk management into how organizations operate (European Data Protection Board, 2021).

The 72-hour breach notification requirement supports containment and transparency. Together, these measures have raised public awareness and improved organizational practices across the EU, even though implementation still varies between sectors and member states (European Commission, 2020).

The U.S. system offers strong protection in narrow areas but lacks overarching safeguards. Every state has a data breach notification law, which creates a baseline of transparency, but thresholds, timelines and enforcement vary widely, and that inconsistency undermines a unified response to threats.

Benefits and drawbacks

GDPR (EU)U.S. sectoral model
BenefitsUser empowerment: strong individual rights build trust and autonomy.
Global influence: a blueprint for laws in other countries, encouraging international cooperation.
Unified standard: one set of rules simplifies compliance for organizations operating across the EU.
Flexibility: rules can be tailored to specific industries.
Lower initial costs: most businesses face fewer compliance costs at the outset.
Innovation-friendly: sectors such as technology have more room to experiment.
DrawbacksCost: broad obligations are expensive, especially for SMEs.
Legal uncertainty: broad scope leaves room for differing interpretations and regulatory risk.
Uneven enforcement: DPAs differ in resources and activity.
Inconsistency: consumers struggle to understand their rights.
Gaps in protection: areas such as ad tech and retail are largely uncovered.
Complex compliance: overlapping multistate requirements can conflict.

At a glance

DimensionGDPR (EU)U.S. sectoral laws
Legal frameworkA single regulationMultiple federal and state laws
CoverageAll sectors and types of personal dataSector-specific
Territorial scopeExtraterritorialMostly domestic
EnforcementNational DPAs coordinated by the EDPBFTC and state attorneys general, decentralized
User rightsExtensive and uniformVary by statute
PenaltiesUp to €20M or 4% of global turnover, whichever is higherVaried, generally lower than in the EU

Conclusion

The GDPR and the U.S. sectoral model sit at opposite ends of the data protection spectrum. The EU's comprehensive, uniform approach has proved effective, but at significant cost. The U.S. model is more flexible and cheaper, and it gives more room for innovation. As technologies such as artificial intelligence increase global data flows, pressure is growing on the United States to adopt a federal privacy law, such as the proposed American Data Privacy and Protection Act (ADPPA), which would harmonize and strengthen protections. My analysis suggests the best path forward is a hybrid model that combines the GDPR's universality with the flexibility and agility of the U.S. approach.

References

  • California Privacy Protection Agency. (2023). CCPA and CPRA overview. https://cppa.ca.gov/laws_regulations/ccpa.html
  • European Commission. (2020). Data protection as a pillar of citizens' empowerment and the EU's approach to the digital transition. https://ec.europa.eu/info/law/law-topic/data-protection
  • European Data Protection Board. (2021). Guidelines on Data Protection Impact Assessment (DPIA). https://edpb.europa.eu
  • European Parliament. (2016). Regulation (EU) 2016/679 (General Data Protection Regulation). https://eur-lex.europa.eu/eli/reg/2016/679/oj
  • International Association of Privacy Professionals. (2024). U.S. state privacy legislation tracker. https://iapp.org/resources/article/us-state-privacy-legislation-tracker/
  • National Conference of State Legislatures. (2024). Security breach notification laws. https://www.ncsl.org/research/telecommunications-and-information-technology/security-breach-notification-laws.aspx
  • Solove, D. J., & Hartzog, W. (2014). The FTC and the new common law of privacy. Columbia Law Review, 114(3), 583–676.