How to read Disclaimer

Part A ยท Foundations

Chapter 1 ยท Orientation: what ethical hackers do

Technical intensity: Low Free preview

Learning outcomes

  1. You can explain what security protects: confidentiality, integrity and availability. You can spot each one in a news story.
  2. You can describe the main security roles: white, grey and black hats; red, blue and purple teams; bug bounty hunters; Security Operations Centre (SOC) analysts; threat-intelligence analysts; digital-forensics specialists. You can say what each role does.
  3. You can list the five stages of a professional engagement: planning and authorisation, reconnaissance, assessment, reporting, remediation and retest. You can say what each stage produces.
  4. You can read a security job advertisement and place it on the career map, with entry routes and certifications for India and other countries.
  5. You can state the one rule that governs every hands-on exercise in this book: never test outside an authorised lab or programme.

Concepts

What security protects: the CIA triad

Almost every security story is about one of three losses. Security professionals call them the CIA triad: confidentiality, integrity and availability. The name has nothing to do with any intelligence agency.

Confidentiality means only permitted people can see information. When a hacker publishes a customer database, confidentiality broke. When a hospital clerk reads a famous patient's records out of curiosity, confidentiality broke. In India, a leak of customer data from a bank or a fintech app is a confidentiality failure. Since the Digital Personal Data Protection (DPDP) Act, 2023 came into force, it can also be an expensive legal failure. Chapter 2 explains this.

Integrity means nobody can change information without permission. Suppose a criminal edits a payment file so that a supplier's invoice pays the criminal instead. Integrity broke. The money moved exactly as the tampered instructions said. Integrity attacks scare finance the most. A ledger that quietly lies is worse than a ledger that is locked.

Availability means the system works when people need it. When ransomware encrypts a hospital's records and ambulances divert to another hospital, availability broke. This is true even if no data was stolen. For a payment network, a stock exchange or a food-delivery app, one hour of downtime means one hour of lost revenue and lost trust.

Most incidents damage more than one part at once. At Equifax, confidentiality broke first: outsiders read private data. Trust broke later: consumers could no longer believe Equifax protected their data. When you read any incident in this book or in the news, name the parts of the triad that were hit. This habit turns a scary story into something you can analyse.

Where the money is: why finance cares most

Security matters everywhere. But finance feels failures fastest, because its product is trusted information moving at speed. India has had clear examples. In August 2018, attackers put malware on the switching system of Cosmos Bank in Pune. They then withdrew cash with cloned cards in thousands of ATM transactions across about 28 countries at nearly the same time. The bank lost โ‚น94 crore (about US$13.5 million).[4] The money was gone in hours. One incident, three failures: stolen card secrets (confidentiality), tampered payment instructions (integrity), and a payment system that had to be partly frozen (availability).

Regulators answered by making security a board-level duty. The Reserve Bank of India (RBI) told banks in June 2016 to adopt a cyber-security framework. It requires a policy approved by the board, a dedicated security operations centre and prompt incident reporting.[5] The Securities and Exchange Board of India (SEBI) combined its earlier circulars into one Cybersecurity and Cyber Resilience Framework in 2024. Its phased deadlines ran into 2025 and cover brokers, depositories, mutual funds and market infrastructure institutions.[6] If you work in banking, fintech or markets, security is not an IT topic. It is a condition of staying in business. This book keeps the finance angle throughout, from crypto forensics in Chapter 29 to cyber insurance in Chapter 32.

Who does what: hats, teams and jobs

You will meet many labels in this field. Here is what they mean, and the one thing you must not get wrong.

White hat, black hat, grey hat. In old Western films, heroes wore white hats and villains wore black hats. Early hacking culture borrowed these words. A white hat tests systems with written permission, inside an agreed scope, to make them stronger. A black hat breaks into systems without permission, usually for money, data or status. A grey hat sits between: probing systems without permission but without clear criminal intent, and sometimes reporting what they find. Now the part that matters for your career and your freedom: grey hat activity is still unauthorised access. Good intentions are not a defence anywhere. In India they are not a defence under the Information Technology Act, 2000. Unauthorised access and damage are punishable whether you meant harm or not. Chapter 2 gives the sections and penalties. The line between white hat and everything else is not intention. It is authorisation.

Red team, blue team, purple team. These words come from military exercises. A red force attacks; a blue force defends. A red team plays the attacker in an approved exercise. It tries to get in the way a real attacker would, under rules agreed in advance. A blue team is the defence: monitoring, detecting, responding, hardening. A purple team is not a third group. It is a way of working: red and blue share their moves while the exercise runs. The defence learns from every attack step immediately, not weeks later in a report. Teams improve fastest when they work this way.

Bug bounty hunters are independent researchers who hunt for flaws in organisations that invite exactly this. The organisation publishes a programme: a public scope, rules, and payment for verified findings. The hunt is lawful because the authorisation came first โ€” the company invited the testing. The moment a hunter steps outside the published scope, the testing becomes unauthorised again. This is why Chapter 2 trains you to read programme rules before touching anything.

SOC analysts work in a Security Operations Centre. This is the place, physical or virtual, where an organisation's logs and alerts are watched around the clock. It is the most common first job in security. The work is detective work: check alerts, decide which are real, escalate the important ones, and write down what happened for the next shift. Threat-intelligence analysts study the attackers themselves: their tools, their infrastructure, their behaviour. They turn this study into warnings the SOC and the business can act on. Chapter 27 does this job against dark-web sources. Digital-forensics and incident-response (DFIR) specialists arrive after something has gone wrong. They preserve evidence, rebuild the intrusion from logs and artefacts, support legal action, and help the organisation recover. In India, DFIR skills matter more and more, next to police cyber-crime units and the forensic auditors who serve banks and regulators.

How an engagement runs: the professional lifecycle

Legal security testing is not a burst of inspired hacking. It is a project with paperwork at both ends. Methods use different words, but the shape is always five stages.

1. Planning and authorisation. Client and testers agree in writing: what will be tested, when, from where, how far each finding may be pursued, and who to call if something breaks or the police arrive. This document โ€” engagement letter, scope, rules of engagement โ€” is what makes everything after it legal. It exists before a single packet is sent.

2. Reconnaissance. The testers collect information about the target, mostly from public sources at first: domain records, certificates, job advertisements, leaked credentials named in breach databases. Chapter 6 covers this. Good recon usually predicts the weak points. Equifax's unpatched internet-facing application is exactly the kind of thing that recon plus a patch check should surface.

3. Assessment. The testing itself: scanning for known weaknesses (Chapters 7โ€“8), probing web applications (Chapters 9โ€“10), controlled exploitation inside the lab or agreed scope (Chapter 12). Testers stop at the agreed boundaries, always.

4. Reporting. Findings are written for two audiences at once. Executives need risk in business terms and a ranked list. Engineers need exact technical detail and fix guidance. Chapter 13 treats report writing as a skill of its own. In professional testing, the report is the deliverable. The testing exists to produce it.

5. Remediation and retest. The client fixes. The testers verify the fixes. A finding closes when the retest confirms it, not when a ticket says so.

Notice that the lifecycle starts and ends with documents and decisions, not commands. Here is the honest answer to "what do ethical hackers do all day": more reading, writing and checking than typing. And every hour traces back to a signed authorisation.

๐Ÿ”ง Technical

In plain words: a scope statement is the written list of what may and may not be attacked. It names exact targets and techniques. Anything not named is off-limits by default.

A workable scope statement answers six questions without vagueness:

  1. Targets: exact IP ranges, hostnames, URLs or applications. In this book, always lab placeholders such as lab.local or 192.0.2.0/24.
  2. Time window: start and end dates, permitted hours, time zone.
  3. Techniques: permitted classes of testing, plus explicit exclusions. For example: no denial-of-service, no physical intrusion, no social engineering of named staff.
  4. Depth limits: how far exploitation may go. Proof of access or full post-exploitation? May production data be touched?
  5. People: named testers with ID numbers, an emergency contact on each side, and who at the client knows the test is running.
  6. Evidence handling: how findings, credentials and any seen data are stored, sent and destroyed.
Example target line (fictional, synthetic):
  in-scope:  https://juice-shop.lab.local  (OWASP Juice Shop VM)
  in-scope:  192.0.2.10-192.0.2.20         (lab host-only range)
  out-of-scope: everything else on Earth, including the office Wi-Fi

The career map, honestly drawn

Security is one of the few technology fields where demand has beaten supply for a decade. The numbers are large. ISC2, the professional body behind the CISSP certification, estimated the global cybersecurity workforce gap at about four million people in its 2023 study. It put India's own gap near 790,000 professionals, about 40 per cent higher than one year before.[7] Such estimates change year to year and method to method. Treat the exact digits as a guide. The direction โ€” many more openings than qualified people โ€” has been stable through the 2020s.

Real entry routes look like this. The most common is the SOC analyst role described above: shift work, alert checks, log reading. The pay is modest at first. The learning is fast, because every shift shows what is attacking organisations right now. A second route runs through system or network administration. People who already run infrastructure move naturally into hardening it, then auditing it. A third route runs through development: programmers move into application security. Chapters 9 and 10 show why that field has more jobs than people. A fourth, newer route is bug bounty. A few independent researchers earn a living from programmes. Treat the success stories with care. Most hunters earn irregularly. The dependable early-career value of bounty hunting is the portfolio of real, documented findings it builds.

Certifications are signals, not skills. But the market reads the signals. At entry level, CompTIA Security+ is the widest recognised ticket. EC-Council's CEH (Certified Ethical Hacker) appears in many Indian job advertisements, partly because government tenders name it. Hands-on mid-level certificates like the eJPT (Junior Penetration Tester) prove practical skill at low cost. The OSCP (Offensive Security Certified Professional) is the respected hands-on benchmark for penetration testers: a 24-hour practical exam that cannot be passed by memorisation. Chapter 33 maps these against career stages. The short version: certifications open doors; labs and portfolios get you hired.

Where are the jobs? Banks, insurers and fintechs in Mumbai, Bengaluru and Hyderabad run large security teams because their regulators demand it. The RBI and SEBI frameworks above create security jobs in effect. Large IT-services firms run security operations for clients worldwide; many Indian analysts start there. Product companies, cloud providers and security-tooling start-ups complete the map. None of this needs you to be a mathematical genius. It needs patience with detail, comfort with being confused while learning, and the discipline to stay inside lines that are written down.

๐Ÿ”ง Technical

In plain words: a vulnerability report is a fixed story: where, what, how bad, how to repeat it, how to fix it. Write it so two engineers who never meet reach the same conclusion.

Every professional report shares this skeleton. Chapter 13 expands it into a full penetration-test report.

  1. Identifier and title โ€” a stable ID and a plain-language title. Example: "Reflected cross-site scripting in search field of juice-shop.lab.local".
  2. Severity with justification โ€” a score (usually CVSS (Common Vulnerability Scoring System); Chapter 8) plus one sentence of business reasoning. The score ranks. The sentence persuades.
  3. Affected asset(s) โ€” the exact host, URL, parameter or version.
  4. Description โ€” what the weakness is, using the OWASP Top 10 or CVE record vocabulary it matches.
  5. Reproduction steps โ€” numbered, from a clean state to the observed impact. The reader must be able to repeat your result.
  6. Evidence โ€” request and response captures, screenshots, log lines. Secrets removed.
  7. Remediation โ€” the specific fix, and how to verify it worked.
Weak:  "XSS found on the search page."
Strong: "Reflected XSS in the q parameter of /search (juice-shop.lab.local);
        payload <script>alert(1)</script> echoes unencoded in 15 lines
        of HTML; CVSS 6.1 (Medium) because session theft needs a
        victim click; fix: context-aware output encoding, see Ch 9."

How this book works

The book has four parts. Part A (Chapters 1โ€“5) builds foundations: law and ethics, networking, Linux, cryptography, and the isolated lab for every hands-on exercise. Part B (Chapters 6โ€“13) is the ethical-hacking sequence: reconnaissance, scanning, vulnerability assessment, web applications, passwords and authentication, exploitation, and reporting. Every technique in Part B is practised only inside your own lab or a legal practice platform. Part C (Chapters 14โ€“31) covers the deep and dark web as a defender's and investigator's subject, and it is the heart of the book: history and honest measurement; the deep web proper; legitimate uses; who uses these spaces and why; how Tor, I2P and the other anonymity networks work; how anonymity fails, technically and operationally; how markets and the underground economy run; how trade moved to messaging platforms; how cryptocurrency moves and hides the money; how investigators took the famous cases apart; how organisations watch all of this as threat intelligence; the law, in India and abroad; and where the field is heading. Part C is analytical throughout. This book never links to illicit services and never instructs anything illegal. Part D (Chapters 32โ€“33) closes with blue-team work, incident response, cyber insurance and a capstone.

Mechanically, every chapter has two reading tracks. Core gives you a complete read. Technical adds the hands-on depth. Switch with the buttons in the page header; the How to read page explains both. Labs come in the same two tiers. The Core lab needs only a browser and a notebook. The Technical lab runs in your own lab. Each chapter ends with a 20-question quiz: 15 core questions and 5 technical questions. The pass mark is 60 per cent. Every answer comes with an explanation. Your progress stays in your browser. Nothing about your reading is sent anywhere.

One rule overrides everything else, repeated in every chapter: every hands-on technique in this book is used only on systems you own or are explicitly authorised to test. The Legal & ethical use page explains why, with the Indian law that enforces it. The next chapter is devoted to exactly this.

๐Ÿ”ง Technical

In plain words: where the security chief sits on the organisation chart tells you whether security can say no to powerful departments. Independence matters more than headcount.

The pattern that works, in line with regulator expectations such as the RBI's board-approved policy[5]:

  • Board / Risk Committee โ€” owns the cyber risk, approves the policy, receives the SOC's numbers.
  • CISO (Chief Information Security Officer) โ€” reports high enough (to the CEO, the risk committee or the board) to stay independent of the teams whose shortcuts they must challenge. "CISO reports to the CIO whose budget they audit" is a known anti-pattern.
  • Security operations (SOC, DFIR) and security engineering sit under the CISO.
  • IT operations and development sit separately. The people who fix findings should not grade their own homework.

When you interview for security roles, ask "where does security report?" The answer tells you whether the organisation wants security or wants the appearance of it.

1 ยท Planning & authorisation 2 ยท Recon- naissance 3 ยท Assess- ment 4 ยท Report- ing 5 ยท Remedi- ation & retest verified fixes feed the next, harder round Figure 1: the engagement lifecycle. It starts with a signature and ends with a verification โ€” not with a keyboard.

Worked demo: one engagement, walked end to end

To see the lifecycle as one continuous story, walk through a small fictional engagement. The company, the testers and every detail are synthetic โ€” invented for this walkthrough. The shape is exactly what a real small web-application test looks like.

The setup. "ApexPay Fintech Labs" (fictional) runs a customer web portal. A competitor's breach is in the news. The board orders an independent security assessment of the portal before the quarterly audit. A two-person team from a fictional firm, "Vetta Security", takes the job.

Stage 1 โ€” planning and authorisation. Vetta and ApexPay sign an engagement letter. It names the single in-scope application (portal.apexpay.lab.local โ€” a placeholder address, as always in this book). It sets a window of ten working days in office hours. It names the two testers, with government ID details. It bans denial-of-service and any touching of the production database contents. It lists an emergency phone contact on each side and encryption rules for all evidence. ApexPay briefs its security lead, the application owner and the local IT staff. The testers get a signed copy of the authorisation to carry. Nothing has been tested yet, and the engagement is already half-defined.

Stage 2 โ€” reconnaissance. The testers map the application's public footprint. They find subdomains, the login and password-reset flows, an old API (application programming interface) documentation page still reachable, JavaScript files that reveal internal endpoint names, and version banners that hint at an outdated support-ticket plugin. Each item is noted as a lead, not yet an attack. Two leads โ€” the stale plugin and the exposed API notes โ€” become the first items of the assessment plan.

Stage 3 โ€” assessment. Working through the agreed plan, the testers confirm three findings. The ticket plugin has a known injection weakness. A broken access-control flaw lets any logged-in customer request other customers' statements by changing an identifier. And the password-reset link is predictable. Exploitation stops at proof, using their own test accounts, per the depth limits. They demonstrate the access-control bug by retrieving a statement that belongs to a second, tester-controlled account. They never touch a real customer's data.

Stage 4 โ€” reporting. The findings are written in the two-voice structure of a professional report. The executive summary ranks the three findings by business risk; the broken access control comes first because it exposes customer financial data. Full technical entries follow, with reproduction steps and fixes. The appendix holds the evidence, with secrets removed. In the findings-review call, ApexPay's engineers reproduce all three findings themselves, from the report alone. That is the practical test of whether a report is any good.

Stage 5 โ€” remediation and retest. ApexPay patches the plugin, adds an object-level access check, and switches reset links to random, unguessable tokens. Eight days later, Vetta retests exactly those three items. The fixes hold. Vetta issues a closure note. The engagement ends not with "we hacked you" but with three signed-off fixes. That is the point of the exercise.

Now hold this walkthrough next to the Equifax story. The difference is not talent or tools. One organisation ran the loop โ€” look at yourself the way an attacker would, write down what you find, fix it, verify the fix. The other did not.

CIA triad
Confidentiality, integrity, availability โ€” the three properties security protects.
White hat / black hat / grey hat
Tester with authorisation; intruder without it; unauthorised prober without clear criminal intent. Grey hat work is still unlawful: the dividing line is authorisation, not intention.
Red team
The attacking side of an approved exercise, copying real attackers under agreed rules.
Blue team
The defending side: monitoring, detection, response, hardening.
Purple team
Red and blue sharing findings in real time, so each attack step teaches the defence at once.
SOC (Security Operations Centre)
The function that watches an organisation's alerts and logs non-stop; the most common entry job in security.
DFIR (digital forensics and incident response)
Preserving and reading evidence after an incident, rebuilding what happened, supporting recovery and legal action.
Threat intelligence
The study of attackers โ€” tools, infrastructure, motives, behaviour โ€” turned into actionable warnings.
Engagement letter
The signed contract that authorises a test and binds both sides.
Scope
The written list of targets, techniques, limits and time window a test may touch. Anything unlisted is off-limits.
Rules of engagement
Operating rules inside the scope: permitted techniques, escalation paths, emergency stops, evidence handling.
Retest
Verification, after fixes, that a finding no longer reproduces. The retest closes the finding.
Bug bounty programme
A public invitation by an organisation to test a defined scope for rewards, under published rules.

Summary

Security protects three things: confidentiality, integrity and availability. Every incident you will read about is a story of at least one of them failing. The Equifax breach showed that the most expensive failures are usually ordinary: a known, patchable flaw that nobody with an attacker's eye was paid to hunt down. The people who do that hunting lawfully are white hats, red teams, bug bounty hunters and penetration testers. Their work is a five-stage lifecycle that starts with a signed authorisation and ends with a verified fix, with writing in both directions. Around them sit the defenders: the SOC that watches, the threat-intelligence analyst who studies attackers, and the DFIR specialist who rebuilds events after a failure. The field is short of people; India's gap alone was near 790,000 professionals in 2023. Finance shows the shortage most clearly, because RBI and SEBI rules make security a condition of doing business. Grey-hat probing is still unlawful; the line that matters is authorisation, not intention. Everything hands-on in this book happens inside your own lab. The next chapter covers the legal ground you will stand on.

Lab โ€” Core tier: the Equifax post-mortem worksheet

โš  Authorized practice only โ€” this exercise, like every exercise in this book, runs only on systems you own or are explicitly authorised to test. Read Legal & ethical use first.

Objective: turn the opening story of this chapter into structured analysis. This habit separates security professionals from spectators. You need only this chapter and a notebook. Every fact you need is cited above.

Steps

  1. Triad. List every harm the Equifax breach caused. Label each harm confidentiality, integrity or availability. Where does "consumers lost trust in a credit bureau" fit? Why is that the hardest one to label?
  2. Failure chain. The chapter lists four failures before the data was taken: the Struts flaw was disclosed with a patch available; the patch was not applied; an internal scan failed to catch it; the inventory of internet-facing systems was weak. Draw them as a chain of arrows. Mark the earliest point where a competent defender could have broken the chain.
  3. Roles. For each role below, write one sentence: how could that role have caught or prevented the breach, had it existed and done its job? SOC analyst; vulnerability-assessment tester; purple team; threat-intelligence analyst.
  4. Lifecycle. Map the incident onto the five stages. Which stages did Equifax effectively skip or fail? What would the reporting stage have produced, and for whom?
  5. Finance angle. The chapter calls a credit bureau a "confidentiality business". In two sentences: why would a bank suffer more from an integrity failure than from an equal-sized confidentiality failure?

Expected result

A one-page worksheet with: (1) at least three harms labelled, and the
trust harm discussed as a second-order confidentiality/availability
effect; (2) a 4-step failure chain whose earliest break-point is "apply
the patch within the window after disclosure"; (3) four role sentences,
each role touching a different link of the chain; (4) lifecycle stages
mapped, noting that assessment and remediation were effectively absent;
(5) an argument that silent ledger tampering persists and compounds,
while a leak is at least discoverable and bounded.

Keep your worksheet. The Chapter 2 core lab continues this story from the legal side.

Lab โ€” Technical tier: is your computer lab-ready?

โš  Authorized practice only โ€” the hands-on tier runs strictly inside your own isolated lab VMs (Chapter 5) or a legal practice platform. Never a real third-party system. Legal & ethical use.

Switch to Technical mode for the hands-on lab.

Objective: check whether your machine can run the Chapter 5 lab, which runs two or three virtual machines at once. Use only the operating system's own built-in tools. These checks only read information. Nothing is installed, changed or removed. The Chapter 5 targets are: CPU with hardware virtualisation enabled; at least 8 GB RAM (16 GB is comfortable); at least 60 GB of free disk.

Steps (pick your operating system)

  1. Windows (10/11): open PowerShell and check memory and virtualisation support:
    Get-CimInstance Win32_ComputerSystem | Select-Object Manufacturer, TotalPhysicalMemory
    systeminfo | Select-String "Hyper-V Requirements"
    Manufacturer      : LENOVO
    TotalPhysicalMemory : 17004224512          (~16 GB โ€” pass)
    
    Hyper-V Requirements:      A hypervisor has been detected. Requirements not applicable
            โ€” OR โ€”
    Virtualization Enabled In Firmware: Yes     (either result is a pass)

    If you see Virtualization Enabled In Firmware: No, enable Intel VT-x or AMD-V in the BIOS/UEFI settings. The Chapter 5 lab guide walks through this safely. For disk space, run Get-PSDrive -Name C and read the Free column, or check the drive properties in File Explorer.

  2. macOS: open Terminal and check CPU, memory and disk:
    sysctl -n machdep.cpu.brand_string
    sysctl -n hw.memsize
    df -h /
    Apple M2                       (Apple Silicon โ€” virtualisation built in; pass)
    17179869184                    (bytes รท 1,073,741,824 โ‰ˆ 16 GB โ€” pass)
    Filesystem  Size  Used Avail Capacity
    /dev/disk3s1 494Gi 312Gi 178Gi   64%  /   (~178 GB free โ€” pass)

    On Intel Macs, VirtualBox handles VT-x on any recent model. Only machines older than about 2010 struggle.

  3. Linux: check virtualisation, memory and disk:
    lscpu | grep -i virtualization
    free -h
    df -h /
    Virtualization:      VT-x
    Virtualization type: full          (pass)
    
                  total    used    free
    Mem:            15Gi     6Gi     8Gi      (pass)
    
    /dev/sda2       234G    171G     51G  78% /  (51 GB free โ€”
                                            below the 60 GB target: free up
                                            space or use an external SSD)
  4. Record your results in three lines: virtualisation yes/no, RAM, free disk. Check them against the targets above. If any line fails, note the fix (BIOS setting, more RAM, external SSD) and set a date to fix it before Chapter 5.

Clean-up

None needed. This lab only reads information and changes nothing.

Quiz

Twenty questions on this chapter: fifteen core and five marked Technical. In Core mode you answer the fifteen. Switch to Technical mode for all twenty. Pass mark 60 per cent. Every answer carries an explanation.

Further reading

  • The House Oversight Committee's Equifax report page โ€” the primary source for this chapter's opening story and findings[1].
  • The FTC's settlement announcement, for the regulatory aftermath and consumer remedies[2].
  • ISC2's annual Cybersecurity Workforce Study โ€” read the latest edition for current workforce numbers[7].
  • Reuters' Cosmos Bank coverage, an early view of a multi-country payment-card fraud[4].
  • The How to read and Legal & ethical use pages of this book, before starting Part A's labs.

References

  1. U.S. House Committee on Oversight and Government Reform, "The Equifax Data Breach" (majority staff report) and press release, 10 December 2018 โ€” oversight.house.gov.
  2. U.S. Federal Trade Commission, "Equifax to Pay $575 to $700 Million as Part of Settlement Related to 2017 Data Breach", 22 July 2019 โ€” ftc.gov.
  3. U.S. Government Accountability Office, Data Protection: Actions Taken by Equifax and Federal Agencies in Response to the 2017 Breach, GAO-18-559, August 2018 โ€” gao.gov.
  4. Reuters, "Suspected cyber criminals steal $13.5 million from India's Cosmos Bank", 14 August 2018 โ€” reuters.com.
  5. Reserve Bank of India, "Cyber Security Framework for Banks", circular DBS.CO/TCSEC/006/31.01.023/2015-16, 2 June 2016 โ€” rbi.org.in.
  6. Business Standard, "Sebi releases new cyber security framework for regulated entities", 20 August 2024 โ€” business-standard.com.
  7. ISC2, Cybersecurity Workforce Study 2023 โ€” isc2.org.
  8. India Code, Information Technology Act, 2000 (as amended) โ€” indiacode.nic.in.

Unverified: check source The 147-million figure for affected Equifax consumers is the FTC's final updated count; earlier official statements said 143โ€“145.9 million. The Cosmos Bank account relies on reporting from the time of an ongoing investigation; final court outcomes, if any, were not available to this book.