What Is Phishing? Examples, Prevention & How to Choose Protection

In the second quarter of 2026, the Anti-Phishing Working Group counted 1,069,681 unique phishing sites. That figure measures attacker infrastructure rather than victims, but it shows the scale of a problem built on one simple idea: persuade a person to hand over access, money, or trust, and the technical defenses never have to be beaten at all. Phishing is the everyday face of social engineering, and learning to recognize it is one of the most useful security skills anyone can have.
TL;DR
Phishing is a deceptive message, call, or web page that impersonates someone you trust to trick you into revealing credentials, approving access, or sending money. It does not need malware to work.
Verify independently. Never use the links, phone numbers, or contact details inside a suspicious message; use a bookmark, an official app, or a number you already trust.
Clicking a link is not automatically a compromise. Entering a password, approving a login prompt, running a file, or sending a payment is far more serious, and each calls for a different response.
Passkeys and FIDO2 security keys resist phishing in a way that typed or texted codes do not, and a password manager adds a useful check because it will not autofill on a lookalike domain.
Businesses need layers: email authentication, filtering, phishing-resistant sign-in, payment controls, easy reporting, and a rehearsed response plan. Application security testing addresses what a stolen login could reach, not the deceptive message itself.
Quick answer: what is phishing?
Phishing is a form of social engineering in which an attacker impersonates a trusted person, brand, or system, using email, text messages, calls, QR codes, or fake web pages, to trick someone into revealing credentials, approving access, installing malware, or sending money. Messages are often urgent and look legitimate, but they do not always contain malicious software.
![]() Protect the Application Behind the Login Phishing-resistant sign-in and staff awareness reduce the chance that a login is stolen. Scoped application testing answers a different question: what could someone do inside your web app, API, or SaaS product if a credential is compromised? Articsledge arranges professionally scoped web application, API, and SaaS application assessments, with scope, authorization, and deliverables agreed in writing first. Scope, personnel, method, deliverables, and timing are agreed before any engagement begins. |
Table of Contents
What Is Phishing?
Phishing is a deception technique in which an attacker pretends to be a legitimate person, company, or system so that the target willingly does something harmful: shares a password, approves a sign-in, opens a file, or sends a payment. The name echoes fishing, with a message as the bait. The MITRE ATT&CK framework catalogs it as technique T1566 and lists attachments, links, third-party services, and voice calls as delivery routes.
How phishing relates to social engineering, spam, spoofing, and fraud
Social engineering is the umbrella term for manipulating people rather than software, and phishing is its remote, message-based branch. Spam is unsolicited bulk messaging and is often just advertising. Spoofing means forging an identity such as a sender address, caller ID, or website, which phishing frequently uses but which also appears elsewhere. Ordinary fraud can happen in person or by post; phishing is fraud that exploits digital channels at scale.
What attackers are after
Credentials and sessions: passwords, one-time codes, or session tokens that unlock email, cloud, and banking accounts.
Money: payments sent to the wrong account, changed supplier details, or diverted payroll.
Data: personal, financial, or company information that can be sold or used for extortion.
A foothold: a file or approval that gives an attacker a way into a device or network, sometimes ending in malware or ransomware.
Permissions: consent for a malicious app to read mail or files without ever learning a password.
Malware is optional. A convincing page that simply records a typed password needs no malicious file, which is one reason scanners alone miss many phishing attempts. Attackers who do gain a foothold may recruit a device into a botnet or install cryptojacking software.
A simple example
Imagine an email from the fictional brand Northwind Storage saying your payment failed and your account will be suspended in 24 hours unless you sign in using the button below. The button opens a page that looks right but sits on an unrelated domain. If you type your password there, the attacker now has it. Nothing was downloaded and nothing was hacked in the usual sense; you were persuaded.
How Does a Phishing Attack Work?
Most attacks follow a recognizable lifecycle. Describing it helps defenders see where to interrupt it, so this overview stays deliberately non-operational.
Pretext and impersonation. The attacker chooses a believable identity, such as a bank, a delivery firm, a manager, or an IT help desk, and a plausible reason to make contact, often drawing on details from public profiles or company websites.
Delivery. The lure arrives by email, SMS, a phone call, a collaboration-app message, a QR code, or even a paid search ad.
Authority or urgency. Deadlines, threats, rewards, or flattery push the target to act before checking.
The requested action. Sign in, call a number, open a document, approve a prompt, scan a code, or change bank details.
Possible compromise. Credentials or session tokens are captured, a payment is redirected, or software is installed, but only if the target complies.
Aftermath. Stolen access is used or sold, and a compromised mailbox may send fresh lures to colleagues and contacts, which is how one success becomes many.
Each step is also a chance to stop the attack: filtering at delivery, a pause at the request, phishing-resistant sign-in at the point of compromise, and fast reporting afterward. Phishing is frequently the opening move of a broader cyberattack, including ransomware incidents, so catching it early limits everything that follows. Broader groundwork, from network security to a sensibly configured firewall, limits what an attacker can do after an initial foothold.
Types of Phishing Attacks
Phishing labels describe overlapping dimensions: who is targeted (everyone or one person), how the lure arrives (the channel), which technique it uses, and what the attacker wants. One incident can be several at once, such as a spear phishing email that leads to a credential-harvesting page and ends in business email compromise.
Type | Channel and targeting | Usual objective | Typical clue |
|---|---|---|---|
Bulk email phishing | Mass email to many recipients | Credentials, payments | Generic greeting, brand and domain mismatch |
Spear phishing | Tailored email or message using personal or role details | Credentials, access, fraud | Accurate context paired with an unexpected request |
Whaling | Aimed at executives or senior staff | Wire transfers, sensitive data | Executive-level urgency and secrecy |
Clone phishing | A copy of a genuine earlier message with a swapped link or file | Credentials, malware | A duplicate of something you already received |
Credential harvesting | A fake sign-in page | Passwords, codes | Web address does not belong to the service |
Business email compromise (BEC) | Email from an impersonated or hijacked executive or vendor account | Payments, payroll redirection | Changed bank details, pressure to skip normal process |
Smishing | SMS or messaging apps | Credentials, payments, malware | Unexpected parcel, toll, or bank text |
Vishing | Phone calls, sometimes with spoofed caller ID | Codes, credentials, transfers | Caller asks for a code or remote access |
Quishing | QR codes in email, posters, or signs | Sends you to a login or payment page | QR code in an unexpected place |
Collaboration and social app phishing | Direct messages in workplace or social apps | Credentials, access | Unfamiliar sender with a file-share lure |
Search or ad-delivered phishing | Paid ads or poisoned search results | Fake login or support pages | Sponsored result with an odd domain |
OAuth (consent) phishing | A real consent screen for a malicious app | Persistent access without a password | Unknown app requesting broad permissions |
Adversary-in-the-middle (AiTM) | A proxy page relaying a real sign-in | Session tokens | A flawless page whose address is wrong |
Voice and video impersonation is getting more convincing, a trend explored in this guide to deepfake detection.
A note on adversary-in-the-middle and consent phishing
In an adversary-in-the-middle attack, a lookalike page relays your real sign-in to the genuine service so the attacker captures the resulting session. In consent phishing there is no fake password box at all: the victim is asked to grant a malicious application access to their account. Both defeat habits built around spotting a fake login form, which is why phishing-resistant authentication and regular reviews of connected apps matter.
Realistic Phishing Examples (Fictional and Harmless)
Every brand, person, and address below is invented for education, and the domains use names reserved for documentation (example.com, example.net, and example.org). These simulations are not working lures.

1. Password reset you did not request
Claim: Northwind Mail says someone tried to sign in from another country. Ask: press Reset password, which points to northwind-secure.example.net. Red flags: an unrequested alert, urgency, and a domain unrelated to the service. Verify: open the app or type the address yourself and check recent activity.
2. Bank verification text
Claim: a text from Harbor Federal says your card is locked. Ask: tap harborfed-verify.example.org and enter card details. Red flags: an unsolicited text with a link and a threat. Verify: call the number printed on your card or use the bank's official app.
3. Parcel or toll payment
Claim: Parcely says a delivery is held until you pay a small fee. Ask: pay at an example.com link. Red flags: a tiny, plausible charge designed to capture card details, and a parcel you may not be expecting. Verify: use the tracking number on the retailer's own website.
4. Payroll change request
Claim: an email appearing to come from an employee asks HR to update bank details before payday. Ask: change the deposit account. Red flags: a free webmail address, urgency, and a bypassed process. Verify: call the employee on a known number and use the documented change procedure.
5. Fake invoice or executive wire request
Claim: a message showing the CEO's name asks finance for an urgent, confidential acquisition payment and says not to call. Ask: wire funds today. Red flags: secrecy, pressure, and a display name that hides a different address. Verify: require a second approver and call back using the company directory.
6. Shared document
Claim: a colleague shared 'Q3 Budget' and you must sign in to view it. Ask: enter work credentials on a page at example.org. Red flags: generic wording and a sign-in prompt on an unfamiliar domain. Verify: ask the colleague through another channel and open files from your own drive.
7. QR code parking payment
Claim: a sticker on a pay station says scan to pay for parking. Ask: scan and enter card details at an example.net page. Red flags: a sticker that looks added on top and a domain that differs from the signage. Verify: use the city's official parking app or pay at the machine.
8. Fake support call
Claim: a caller from Contoso Cloud Support says your account has been breached. Ask: read out a verification code or install a remote-access tool. Red flags: you did not initiate contact, and the caller wants a secret. Verify: hang up and call the number on the company's official website.
9. Recruitment lure
Claim: a recruiter from Fabrikam Talent offers a remote role after a quick chat. Ask: submit ID scans and bank details for a background check. Red flags: an unsolicited offer, early requests for sensitive data, and a push to a private messaging app. Verify: contact the company through its careers page.
10. Suspicious cloud sign-in
Claim: a New sign-in alert from Contoso Cloud, followed by a stream of approval prompts on your phone. Ask: approve a prompt or press Secure my account. Red flags: prompts you did not trigger and a link in the alert. Verify: deny the prompts, then go to the service directly, review sign-in activity, and change your password.
How to Recognize Phishing: Warning Signs and a Checklist
No single clue is conclusive. Poor grammar, a padlock icon, a familiar logo, or an internal-looking sender can each appear in genuine and malicious messages alike, so treat clues as reasons to verify rather than as proof either way. The strongest habit is independent verification: confirm through a channel you already trust.
Phishing Identification Matrix
This matrix is Articsledge editorial synthesis drawn from the public guidance in the references. It is not the result of testing or a survey, and it is meant as a starting point rather than a complete list.
Pattern | Requested action | Warning sign | Safe independent check |
|---|---|---|---|
Account alert email | Click to sign in or reset a password | Unrequested alert, link domain differs from the service | Open the app or type the address yourself |
Parcel or toll text | Pay a small fee through a link | Unexpected sender, odd domain | Check the carrier or toll operator directly |
Executive or vendor payment request | Wire money or change bank details | Secrecy, urgency, changed details | Call back on a known number; require a second approval |
Shared document | Sign in to view | Unexpected share, generic wording | Confirm with the sender on another channel |
QR code | Scan, then pay or log in | Sticker over a code, mismatched domain | Use the official app or website |
Phone call from support or a bank | Share a code or install software | The caller started contact and asks for secrets | Hang up and call the official number |
App permission screen | Grant access to your account | Unknown app, broad permissions | Check the publisher, decline, and review connected apps |
A practical checklist
Did I expect this message, call, or request?
Does the real sender address or number match the organization, not just the display name?
Does the web address belong to the service when I read it carefully, especially the part before the first single slash?
Is it pushing urgency, secrecy, or fear?
Is it asking for a password, code, payment, or permission in a way I would not normally provide it?
Did a file or QR code arrive unexpectedly?
Can I confirm through a bookmark, an official app, or a phone number I already trust?
If I am unsure, can I pause and report it instead of acting?
Why Phishing Works
Phishing succeeds because it targets ordinary human strengths such as trust, helpfulness, and speed. It is not a character flaw.
Context fits. Messages arrive during real workflows, like invoices at month end or parcels near the holidays.
Authority and urgency shorten the pause in which doubt normally forms.
Small screens and busy moments hide detail, and mobile layouts often truncate addresses.
Attackers iterate. They test lures at volume and keep whatever works, while defenders have to catch every attempt.
Organizational factors matter just as much: unclear payment approval rules, passwords reused across services, shared inboxes, and a culture in which reporting a mistake feels risky. Blaming individuals misses the point. Good systems assume that someone will eventually click and make that click survivable. For a wider view of who is behind these campaigns, see this overview of threat actors.
What Can Happen If a Phishing Attack Succeeds?
Keep attempts and compromises separate. An attempt is a message that was sent; a compromise happens only if the target complies. When one does, the outcomes range from minor to severe. For how attackers think more broadly, see this guide to hacking.
Account takeover and session exposure: an attacker signs in as you, or reuses a stolen session token without needing your password again.
Identity theft and data exposure: personal or company information is copied, sold, or used for extortion.
Unauthorized payments: funds are sent to attacker-controlled accounts, or supplier and payroll details are quietly changed.
Malware: an opened file can install tools such as a keylogger or a loader for later attacks.
Operational and reputational harm: downtime, recovery costs, notification duties, and lost customer trust.
Three documented cases
Twitter, July 2020. Attackers phoned employees while posing as IT staff and sent them to a fake VPN sign-in page. According to the New York Department of Financial Services investigation, 130 accounts were targeted and tweets were sent from 45 of them in a cryptocurrency scam. The lesson: voice phishing combined with a convincing page can beat employees even at a technology company.
Facebook and Google invoice fraud, 2013 to 2015. A Lithuanian fraudster impersonated the real Taiwan-based supplier Quanta Computer and sent forged invoices and contracts to staff at both companies. Court-related reporting puts the amounts at roughly $23 million from Google and about $98 million from Facebook, and he received a five-year prison sentence in 2019. The lesson: payment fraud can bypass technology entirely, so approval and verification processes are the real defense.
Uber, September 2022. Uber reported that an attacker likely bought a contractor's password and then repeatedly triggered two-factor approval requests until the contractor accepted one. The attacker then reached other employee accounts and tools, including Slack and G-Suite. The lesson: approval prompts can be worn down, which is why phishing-resistant sign-in is stronger than push approvals.
![]() A Stolen Login Shouldn't Expose Every Account When a credential is stolen, the next question is how far it can reach. Well-designed access control, session handling, authorization, and tenant isolation keep one compromised account from becoming a wider breach. Scoped application and API security testing examines those boundaries in your own product, within a scope and authorization you agree in writing. Testing does not stop phishing messages; it examines what a compromised account could reach. |
Phishing Statistics and Recent Trends
Phishing statistics are easy to misuse because sources count different things. The table ties each figure to its source, period, and definition, and the rows should not be added together or treated as measuring the same thing.
Source | Period | What it measures | Figure | Interpretation and limits |
|---|---|---|---|---|
April to June 2026 | Unique phishing sites (unique bases of URLs found in reported phishing emails) | 1,069,681, up 10.1% from 971,181 in Q1 2026 | Counts attacker infrastructure, not victims; one site can be advertised through thousands of URLs. | |
APWG Q2 2026 | Q2 2026 | Share of attacks by targeted sector | SaaS and webmail 29.1%; payment 13.2%; financial institutions 11.4% | Share of attacks by brand sector, not a measure of losses. |
APWG Q2 2026 (Fortra BEC data) | Q2 2026 | Wire-transfer BEC requests | Average request $61,732, up from $42,663 | An average from one data partner, not total losses. |
Calendar 2025 | Public complaints in the Phishing/Spoofing category | 191,561 complaints; $215,843,126 in reported losses | Reported complaints only; many victims never report, and incidents may be filed under other categories. | |
FBI IC3 2025 | Calendar 2025 | Business Email Compromise complaints | 24,768 complaints; $3,046,598,558 in reported losses | A separate category from Phishing/Spoofing, so the two are not directly comparable. |
FBI IC3 2025 | Calendar 2025 | All complaints received | 1,008,597 complaints; $20.877 billion in reported losses | Total internet-crime losses, most of which are not phishing. |
Read together, the sources show that phishing is high-volume and shifts targets quickly. APWG reported that smishing rose 40% and vishing 20% from the first to the second quarter of 2026. Losses reported to the FBI under business email compromise were far larger than those filed under phishing and spoofing, partly because the categories differ. None of these figures can tell you your own odds of being targeted, and detection-focused tools such as AI fraud detection address only part of the picture.
How Individuals Can Prevent Phishing
Start with the few actions that give the most protection, then add optional precautions as time allows.
The strongest actions
Verify independently. Do not use links, numbers, or QR codes in the message. Use a bookmark, an official app, or a number from your card or the company's website.
Use passkeys or security keys where offered. Properly implemented FIDO2 passkeys are bound to the genuine website, so a lookalike page cannot use them. NIST's authentication guidance treats manually entered codes as not phishing-resistant.
Use a password manager with unique passwords. It will not autofill on a lookalike domain, and unique passwords keep one theft from unlocking everything.
Turn on the strongest multi-factor option available. Treat an unexpected approval prompt as a sign your password may be known: deny it and change the password.
Protect account recovery. Keep recovery email addresses and phone numbers current and secured, because attackers often go after the recovery path.
Helpful additional precautions
Keep your operating system, browser, and apps updated.
Turn on sign-in and payment alerts, and consider virtual or single-use cards for online purchases.
Review connected apps and email forwarding rules a couple of times a year.
Report phishing to your email provider and to your local authority; in the United States, the FTC's phishing guidance explains how.
How Businesses Can Reduce Phishing Risk
No single control stops phishing. Layer them so one failure does not become an incident, and be clear about what each one does and cannot do. Teams weighing automation can start with this overview of AI in cybersecurity.
Email authentication (SPF, DKIM, DMARC): makes it harder for others to forge your exact domain. It does not stop lookalike domains, compromised accounts, or attackers sending from their own correctly configured domains.
Filtering and sandboxing: catches known-bad and suspicious messages, though brand-new campaigns can slip through.
Identity controls and phishing-resistant MFA: make a stolen password insufficient. Passkeys and FIDO2 keys resist phishing; push approvals and typed codes are weaker.
Least privilege: limits what any single compromised account can reach.
Payment controls: call-back verification and dual approval for bank-detail changes stop business email compromise where technology cannot.
Endpoint, browser, and DNS protection: blocks known malicious sites and payloads but cannot judge a request for a wire transfer.
Training and easy reporting: builds the habit of pausing and reporting. It supports, and cannot replace, technical controls, and quality varies across phishing simulation software.
Monitoring and response: spots abnormal sign-ins, new mailbox rules, and token reuse; see this guide to threat detection and response.
SaaS, application, and vendor security: reviews third-party app permissions and supplier access, which connects to wider cloud security practice.
Recovery drills: rehearse the response and verify that backups, such as with Backup as a Service, actually restore.
7/30/90-day Business Defense Roadmap
This roadmap is Articsledge editorial synthesis. It is not drawn from testing, and roles and timing should be adapted to your size, tools, and risk.
By | Priority action | Accountable role | Dependencies | Evidence of completion |
|---|---|---|---|---|
Day 7 | Enforce MFA on email and admin accounts; set up a simple way to report suspicious messages | IT lead | Identity provider access | MFA coverage report; a tested reporting route |
Day 7 | Require call-back and two approvals for payment and bank-detail changes | Finance lead | A written policy | Approved procedure and a sample audit |
Day 30 | Publish SPF, DKIM, and DMARC in monitoring mode and review mail flow | IT lead | DNS access and a list of legitimate senders | Records published; reports reviewed |
Day 30 | Inventory connected apps and forwarding rules and remove unknown ones | IT or security lead | Admin console access | Inventory and removals logged |
Day 30 | Run a short role-based awareness session with a reporting walkthrough | Security or HR lead | Reporting route live | Attendance and first reports received |
Day 90 | Move administrators and high-risk staff to passkeys or security keys | IT lead | Supported devices and a recovery process | Enrollment rate; recovery tested |
Day 90 | Move DMARC toward enforcement | IT lead | A clean sender inventory | Policy change recorded |
Day 90 | Rehearse a tabletop covering a phished mailbox and a fraudulent payment | Security lead with finance | A response plan and contact list | Written lessons and follow-up actions |
Day 90 | Plan scoped application and API testing for customer-facing systems, if you run them | Engineering lead | Written authorization and agreed scope | Signed scope; report and retest plan |
What to Do If You Clicked, Entered Data, or Paid
Clicking alone does not always mean compromise. Stay calm, work from a trusted device, and match your response to what actually happened. Reporting channels and legal duties differ by country and sector, so treat these steps as general guidance and adapt them locally.
Phishing Incident-Response Table
What happened | Do first | Then | Escalate when |
|---|---|---|---|
You received a suspicious message | Do not click, reply, or open attachments | Report it using your email provider's report option, then delete it | You interacted with it in any way |
You opened a link but entered nothing | Close the page | Watch for unexpected downloads or prompts and keep your device updated | A file downloaded or a permission prompt appeared |
You entered a password | Change it on the genuine site from a trusted device, and anywhere you reused it | Sign out other sessions and check recovery settings, forwarding rules, and connected apps | It was a work or financial account |
You entered a one-time code | Treat the account as compromised and change the password | Revoke sessions and move to a passkey or security key if available | The account holds sensitive data |
You approved a login prompt you did not start | Deny future prompts, sign out everywhere, and change the password | Review sign-in history and re-register your MFA methods | It is a work account |
You opened or ran a file | Disconnect the device from the network | Scan with a trusted tool and change passwords from a different device | A work device is involved or symptoms persist |
You entered card or payment details | Call your card issuer using the number on the card | Freeze or replace the card and review statements | Charges you do not recognize appear |
You sent money | Contact your bank immediately and ask about recall options | Keep emails, receipts, and account details as evidence | Always report it; in the US, the FBI's IC3 accepts complaints |
You granted OAuth consent to an app | Revoke the app in your account's connected-apps settings | Change the password and check for mailbox rules | Work data was reachable by the app |
You suspect a work account is compromised | Tell your IT or security team right away | Preserve messages and screenshots; administrators reset sessions and review logs | Customer or regulated data may be exposed, since notification duties vary by jurisdiction |
Malicious activity does not always leave an obvious file, as the guide to fileless malware explains, so unexplained behavior after a click deserves attention even when a scan comes back clean. Automated AI threat detection can flag unusual sign-ins or device behavior that a person might miss.
Which Technologies Protect Against Phishing?
Tools play three roles: stopping or detecting the lure, making stolen credentials less useful, and testing weaknesses that would worsen the consequences. This comparison deliberately names no product rankings.
Control | Directly stops or detects phishing? | Strengths | Limits |
|---|---|---|---|
Native email security | Partly | Built into the mail platform; filters common spam and known threats | Fresh, targeted lures can pass |
Secure email gateways | Yes | Deeper link, attachment, and impersonation checks | Needs tuning and cannot judge a payment request |
Web, DNS, and browser filtering | Yes | Blocks known malicious sites at click time | Misses new sites and does nothing for phone fraud |
Endpoint protection and EDR | Partly | Detects malicious behavior after a file runs; see endpoint protection | Does not stop a password typed into a fake page |
Password managers | Indirectly | Will not autofill on mismatched domains; encourages unique passwords | Users can still paste credentials manually |
Passkeys and security keys | Reduces impact | Bound to the genuine site, so lookalikes cannot capture them | Needs service support and a recovery plan |
Identity protections and conditional access | Reduces impact | Risk-based sign-in and session controls | Needs correct configuration and administration |
Training and simulation | Builds behavior | Improves reporting habits | Not sufficient alone; results vary |
Monitoring and managed security | Detects aftermath | Spots abnormal activity, as in SIEM tools | Needs people and process; does not prevent the lure |
Authorized security assessments | No | Examine application weaknesses that would worsen stolen-credential consequences | Do not stop or simulate phishing unless separately scoped |
Attackers and defenders are both adopting automation, a shift covered in this overview of AI cybersecurity tools. Phishing is also a delivery route for AI-powered malware and AI ransomware, so detection at the endpoint still matters.
How to Choose the Right Phishing Protection
Begin with your own situation: which channels you actually use, what identity and email platforms you run, which controls you already have, and where the gaps are. Then compare options on evidence rather than marketing.
Articsledge Phishing Protection Decision Matrix
This is an illustrative 100-point editorial rubric. Score each criterion from zero up to its maximum using evidence from trials, documentation, and independent sources. The weights are a judgment, not validated research, and the matrix makes no claims about any specific product.
Criterion | Maximum points | How to score |
|---|---|---|
Coverage of your real exposure (email, SMS, voice, SaaS apps) | 20 | Map the channels you use; award full points only if the option covers them |
Fit with your identity and email stack | 15 | Documented and tested integration |
Effect on stolen credentials (phishing-resistant sign-in, session controls) | 15 | Passkey and FIDO2 support, session revocation |
User experience and accessibility | 10 | Friction, clarity of warnings, accessibility support |
Privacy and data handling | 10 | What data leaves your environment and under what terms |
Deployment and administration effort | 10 | Time, skills, and ongoing maintenance |
Evidence quality and honesty of claims | 10 | Independent evidence and stated limits |
Support and total cost | 10 | Quality of support, licences, training, and hidden costs |
Add the scores. Roughly 80 or more suggests a strong fit, 60 to 79 a workable fit with identifiable gaps, and below 60 a reason to keep looking. Those bands are a starting convention, not a validated standard.
Scenario guidance
Who you are | Sensible starting point | Usually fine to defer |
|---|---|---|
Individual | Password manager, passkeys, verification habits | Paid security suites unless you need a specific feature |
Freelancer | Phishing-resistant MFA on client tools, payment verification, tuned email security | Enterprise gateways |
Small firm | Tuned built-in email security, DMARC, security keys for administrators, payment controls | A dedicated security operations team |
SaaS team | Conditional access, session hygiene, and scoped application testing | Broad tooling that duplicates your identity provider |
Enterprise | Layered stack, monitoring, response runbooks, regular testing | Single-vendor assumptions |
Red flags in vendor claims
Promises of complete or guaranteed protection.
No description of what the product cannot catch.
Missing integration documentation for your platform.
Pressure to sign quickly, or vague answers on data handling.
If you lack in-house expertise, these guides to cybersecurity services and cybersecurity consulting services explain what to ask for and when outside help makes sense. Rules such as the NIS2 Directive and wider cloud compliance duties can also shape what protection you need.
![]() Validate the Security Layer Your Tools Don't Assess Email filters, passkeys, and endpoint tools address the lure and the login. They do not tell you whether your own web application, API, or SaaS product enforces access rules correctly once someone is signed in. When a focused review makes sense, Articsledge can discuss a scoped assessment, with written authorization and deliverables agreed first. Availability and scope are confirmed before any commitment, and no outcome is guaranteed. |
When Can Penetration Testing Help With Phishing-Related Risk?
Penetration testing helps with the consequences of phishing, not with the deceptive message itself. If an attacker obtains a valid login, the next question is what that login can do. A scoped assessment of a web application, API, or SaaS product examines authentication, authorization, session handling, password reset and account recovery, API tokens, role boundaries, tenant isolation, and sensitive business workflows, which are the places where one stolen credential can become a wider incident. For background on the discipline, see this penetration testing services guide and the explainer on vulnerability assessment. Because an API often exposes the same data as the user interface, and SaaS platforms host many customers together, both deserve the same scrutiny as the web front end.
What a sound engagement looks like
Written authorization and scope: the systems, environments, and test accounts that are in scope, plus the rules of engagement.
Qualified delivery: scope, personnel, method, deliverables, schedule, and commercial terms are agreed first, and availability depends on the specialists engaged.
Useful reporting: prioritized findings, remediation guidance, and optional retesting to confirm that fixes work.
What it does not do
A penetration test does not stop or simulate phishing unless that is separately scoped. It does not replace email filtering, phishing-resistant sign-in, staff training, or incident response, it cannot guarantee that every weakness will be found, and it is not a compliance certificate. Its value lies in answering a narrower question honestly. Automation is changing parts of this field, as discussed in AI penetration testing.
Articsledge arranges professionally scoped web application, API, and SaaS application assessments, and related vulnerability assessment and remediation verification can be discussed subject to scope and specialist availability. Someone recovering a hacked personal account does not need this; the response table above is the better place to start.
![]() Request a Clearly Scoped Web App or API Assessment If you run a customer-facing web application, API, or SaaS product and want to understand what a compromised account could reach, start with scope. Articsledge can discuss written scope, authorization, and deliverables with business buyers, and confirm whether qualified delivery is available before anything is agreed. No pricing, timelines, or results are promised until scope is agreed in writing. |
Phishing Myths and Misconceptions
Myth: HTTPS proves a site is legitimate. HTTPS means the connection is encrypted, not that the owner is trustworthy, and lookalike domains can have valid certificates too.
Myth: all MFA is phishing-resistant. Typed codes and push approvals can be captured or worn down; FIDO2 passkeys and security keys are designed to resist phishing.
Myth: security software blocks every scam. Filters miss new lures, and voice, text, and payment fraud often bypass them entirely.
Myth: small companies are not targets. Attacks run at volume, and smaller firms often have fewer controls and attractive payment processes.
Myth: every link click infects a phone. Usually the risk is what you enter or approve afterward, although unexpected downloads or prompts deserve attention.
Myth: email authentication makes a message trustworthy. SPF, DKIM, and DMARC reduce exact-domain forgery, but compromised accounts and lookalike domains can still pass.
Myth: penetration testing stops phishing. It examines applications and systems; it does not filter messages or train people.
Frequently Asked Questions About Phishing
What is phishing in simple terms?
Phishing is a scam in which someone pretends to be a trusted person or organization to get you to share a password, approve a login, open a file, or send money. It usually arrives as an email, text, call, or QR code, and it does not always involve malware.
What is the difference between phishing, smishing, and vishing?
Phishing is the general term for impersonation-based lures, most often delivered by email. Smishing uses text messages or messaging apps, while vishing uses phone calls or voice messages. The channel changes, but the goal is the same: persuade you to act before you verify through a trusted route.
How can I tell if an email is phishing?
Check whether you expected it, whether the real sender address and link domain match the organization, and whether it pushes urgency or secrecy. No single clue is conclusive, so verify through a bookmark, official app, or a phone number you already trust rather than the message itself.
What should I do if I clicked a phishing link?
Close the page and do not enter anything. Clicking alone does not always mean compromise. Watch for unexpected downloads or prompts, keep your device updated, and tell your IT team if it was a work device. If you entered data or approved a prompt, follow the matching row in the response table above.
What if I entered my password on a fake page?
Change the password on the genuine site from a trusted device, and change it anywhere you reused it. Sign out other sessions, check recovery settings, forwarding rules, and connected apps, and turn on the strongest sign-in method available. If it is a work account, notify your administrator immediately.
Are one-time codes as safe as passkeys?
No. A typed or texted code can be entered on a fake page and relayed to the real service, so NIST does not treat manually entered codes as phishing-resistant. Properly implemented passkeys and FIDO2 security keys are tied to the genuine website and are designed to resist that.
Can phishing happen without malware?
Yes. Many attacks only need you to type a password, read out a code, approve a prompt, grant app permissions, or send a payment. Because no malicious file is involved, antivirus tools may see nothing wrong, which is why verification habits and strong sign-in methods matter.
Does a padlock or HTTPS mean a website is safe?
No. The padlock only shows that the connection is encrypted, not that the site is run by who you think. Attackers can obtain certificates for lookalike domains, so read the address carefully and reach important accounts through a bookmark or official app instead of a link in a message.
What is business email compromise?
Business email compromise is fraud in which an attacker uses a spoofed or hijacked email account, often posing as an executive or supplier, to redirect payments or sensitive data. It can involve no malicious link at all, which is why call-back checks and dual approval for payment changes are key defenses.
How can a small business protect itself from phishing?
Start with phishing-resistant or at least strong multi-factor sign-in on email and admin accounts, call-back and dual approval for payment changes, SPF, DKIM, and DMARC, a simple way to report suspicious messages, and a rehearsed response plan. Add tools as your risk and budget grow.
Do I need paid anti-phishing software?
Not always. Individuals can get far with a password manager, passkeys, updates, and good habits, and many email platforms include filtering. Paid tools make sense when you have specific gaps, such as targeted impersonation or complex integrations. Evaluate them on evidence and fit, not marketing claims.
How much does phishing protection cost?
Costs range from free built-in features to subscription tools and managed services, and depend on your size, platforms, and support needs. Compare total cost, including administration time, training, and integration work, rather than license price alone, and be wary of any vendor promising complete protection.
Does penetration testing prevent phishing?
No. Penetration testing examines applications and systems for weaknesses and does not filter messages, train staff, or stop deceptive lures unless a phishing exercise is separately scoped. It can help limit the damage a stolen login could cause inside a web app, API, or SaaS product.
How do I report phishing?
Use the report option in your email app, forward scams to your provider or employer's security team, and report to your national or local authority. In the United States, the FTC and the FBI's IC3 accept reports. If you lost money, contact your bank first.
What is quishing?
Quishing is phishing delivered through a QR code, for example in an unexpected email, on a poster, or on a sticker placed over a parking sign. Because a phone hides the full address until you scan, check where the code leads and use the official app or website instead when in doubt.
Key Takeaways
Phishing is deception, not just malware: a password typed into a fake page is enough.
Verify through a bookmark, an official app, or a number you already trust, never through the suspicious message itself.
Clicking is not always a compromise; entering data, approving prompts, running files, or paying is.
Passkeys and FIDO2 security keys resist phishing, while typed or texted codes are weaker.
Business email compromise is largely a process problem, and call-backs plus dual approval work.
Layer filtering, authentication, identity controls, training, monitoring, and response, and know the limits of each.
Statistics measure different things, so do not equate phishing sites, complaints, and losses.
Penetration testing examines what a stolen login could reach; it does not stop phishing.
Actionable Next Steps
Individuals, today (free): set up a password manager and replace reused passwords, starting with email and banking.
Individuals, today (free): turn on passkeys or a security key where offered, and review account recovery options.
Individuals, this week: save your bank's official number and adopt one rule: never act on contact details in an unexpected message.
Businesses, this week: enforce multi-factor sign-in on email and admin accounts, and require call-back and dual approval for payment-detail changes.
Businesses, this month: publish SPF, DKIM, and DMARC in monitoring mode and give staff a simple way to report suspicious messages.
Businesses, this month: write a one-page response plan from the table above and rehearse it. For a broader foundation, read what cybersecurity is.
Before buying anything: score your options with the decision matrix and fill gaps in order of risk.
If you run a web app, API, or SaaS product: consider scoped penetration testing to learn what a compromised account could reach.
Glossary
Phishing: deceptive messages, calls, or pages that impersonate a trusted party to obtain access, data, or money.
Spoofing: forging an identity such as a sender address, caller ID, or website.
BEC (business email compromise): fraud using spoofed or hijacked business email to redirect payments or data.
Smishing: phishing by text message.
Vishing: phishing by voice call.
Quishing: phishing through a QR code.
MFA (multi-factor authentication): requiring more than a password to sign in.
Passkey: a phishing-resistant credential bound to a genuine website and unlocked on your device.
FIDO2: an open standard behind passkeys and hardware security keys.
SPF: a DNS record listing servers allowed to send mail for a domain.
DKIM: a cryptographic signature that helps verify a message was not altered.
DMARC: a policy that tells receivers how to handle mail failing SPF and DKIM checks.
IAM (identity and access management): the tools and policies that govern who can access what.
OAuth: a standard that lets an app access an account without the password.
Session token: a credential proving you are already signed in.
EDR (endpoint detection and response): software that monitors devices for malicious behavior.
Penetration testing: an authorized, scoped attempt to find exploitable weaknesses.
Sources & References
APWG Phishing Activity Trends Report, Q2 2026, Anti-Phishing Working Group, published 28 August 2026.
2025 Internet Crime Report, FBI Internet Crime Complaint Center (IC3), covering 2025 data; the document does not state a publication date.
Phishing, Technique T1566, MITRE ATT&CK, MITRE Corporation.
NIST SP 800-63B-4, Digital Identity Guidelines: Authentication and Authenticator Management, National Institute of Standards and Technology.
How to Recognize and Avoid Phishing Scams, Federal Trade Commission, Consumer Advice.
Twitter Investigation Report: A Report on Investigation of Twitter's July 15, 2020 Cybersecurity Incident, New York State Department of Financial Services, issued October 2020.
Security update, Uber Newsroom, 16 September 2022, updated 19 September 2022.
Man pleads guilty to phishing scheme that fleeced Facebook, Google of $100 million, WGBH, 25 March 2019.
Five years in the clink for super-crook who scammed Google, Facebook out of $120m with fake tech invoices, The Register, 20 December 2019.






