Which of these is not an example of a possible tailgating attack? A request that does not involve an unauthorized person following someone through a physically secured entry point is not a tailgating attack. Tailgating is a physical intrusion technique: an attacker walks in behind an authorized person, or persuades them to hold a door, to reach a restricted area without credentials. Anything that does not model a human passing through a controlled portal is a different attack class, including phishing, malware delivery, and network eavesdropping.
The confusion is common because "tailgating" appears in both physical security and network security literature. In networking, "tailgating" can describe following another data frame into a switched segment, or VLAN hopping. Both senses describe following an authorized entity. Neither describes a credential-corrupted login, a phishing email, or a brute-force password guess.
Definition mechanics and why following works
A tailgating attack exploits a design assumption in access control: that one successful authentication equals one entering person. The mechanism is simple. An authorized badge holder presents a credential, the reader unlocks the door, and a second person passes through the same opening before it closes.
The attack works because door hardware enforces a state machine with one residual permission state. The "door open" state typically lasts a few seconds, called the held-open time or unlock delay. Commercially, that delay runs roughly 5 to 30 seconds, and it exists to allow handicapped access and legitimate multi-person passage. That window is the entire attack surface.
Tailgating is passive and low-skill. It requires no forged badge, no cloned credential, and no software. It exploits politeness and social norms. A person carrying boxes, wearing a uniform, or holding a badge-shaped lanyard creates a plausible reason to be admitted. This is why physical tailgating sits in the same attack family as piggybacking: the difference is consent. Piggybacking happens with the authorized person's knowledge; tailgating happens without it.
The counterintuitive detail: the most effective tailgating defense is not a better lock. It is a norm that makes holding the door socially awkward. Security consultants consistently find that the human courtesy of door-holding defeats hardware that costs thousands of dollars per portal.
Quantified context and measured exposure
Physical tailgating is well documented in insider-threat and red-team literature, though exact prevalence numbers are scarce because successful tailgating is often never logged. Access-control systems record valid badge reads, not the number of bodies that followed. That logging blind spot is itself the key finding.
Red-team engagements routinely report that tailgating succeeds at a substantial share of entry attempts in office environments, often reaching well above half of attempts at main lobbies and pedestrian entrances. These figures come from private assessment reports and vary enormously by building. Treat any single percentage as site-specific rather than universal.
The dollar stakes are easier to size. A single access-control door with card reader, lock, and controller commonly costs in the range of several thousand US dollars per opening to install, before maintenance. Pedestrian turnstiles or optical mantrap portals run much higher, often tens of thousands of USD per lane. Yet a tailgate at a lobby cost center bypasses all of it in the seconds it takes to say "I forgot my badge."
Time matters too. Badge cloning, by contrast, requires physical proximity for an RFID read, sometimes seconds of card handling. Tailgating requires zero interaction with the card. It removes the electronics problem entirely and reintroduces the social-engineering problem.
Edge cases and the network double meaning
The security term has a second life in networking. In a switched LAN, tailgating can describe an attacker following an authorized device onto a port that is already authenticated, such as via 802.1X port sharing or MAC-based authentication caches. The abstract structure is identical: an authorized entity opens a gate, and an unauthorized one rides through.
There are other edge cases that trip people up. A person who uses a lost and found badge is not tailgating; that is credential misuse. An authorized employee who deliberately lets a contractor in for a valid delivery is not tailgating; that is a policy bypass or an unauthorized escort, depending on consent. A vehicle barrier lifted by a driver who follows a delivery truck is a tailgating variant, a "follow-in," often called vehicle tailgating.
Warning signs of a possible tailgating attack:
Someone waits near a door without swiping a badge, holding a plausible prop or story. A person carries a box, coffee tray, or folded lanyard to justify being admitted. Someone asks badge holders for a door hold and pivots toward a restricted zone. A group that splits up, with each member entering through a different door. An unbadged person in a restricted area who claims to be "with IT" or "waiting for someone."
The list above describes people following through doors. Anything else on a generic security quiz that involves no portal following is not tailgating.
Common misconceptions
The most common misconception is that tailgating requires some technical skill with badges or locks. It requires none. That is precisely why it persists alongside 2010s-era electronic access controls.
The second misconception is terminology fragmentation. Many incident-response quizzes list phishing email delivery, USB drop attacks, shoulder surfing, and brute-force login attempts as candidate attacks. Only shoulder surfing is in the same family as tailgating, because it follows an authorized person's actions to capture what they typed. Phishing and credential guessing are not following a person through a control; they are coercion or corruption of the authentication itself.
A third misconception is that tailgating only happens inside corporate offices. It occurs at data centers, hospitals, airports, schools, warehouses, and research labs, and it extends to vehicles and to network ports. The shared feature is an authenticated passage with an open interval.
A fourth misconception is that logged badge reads prove nobody tailgated. Access logs can show one valid read and two entrants. The log cannot distinguish them. Audits that count badge events therefore underestimate physical entry counts.
A professional caveat: some vendors and internal policies lump "tailgating" and "piggybacking" together, while forensic reports sometimes call both "unauthorized accompaniment." When reading a quiz answer key, follow the mechanism, not the label.
Which items are not tailgating attacks
The phrase "which of these is not an example of a possible tailgating attack" usually comes from a training quiz. The correct answer is the option that does not involve following an authorized entity through a physical or logical gate. To make the boundary concrete, compare attack categories side by side.
| Attack type | What is followed or exploited | Physical portal? | Requires social consent? | Belongs to tailgating family? |
|---|---|---|---|---|
| Door following (classic) | A badge holder passing a door | Yes | Sometimes | Yes |
| Vehicle follow-in | A car entering a barrier gate | Yes | Rarely | Yes |
| Shoulder surfing | A person's screen or keypad entry | Rarely | No | Adjacent family |
| 802.1X port following | An authenticated network device | No | No | Network variant |
| Phishing email | A user's trust, not a gate | No | Yes (deceit) | No |
| Badge cloning | Credential cryptography | Yes, entry only | No | No |
| Brute-force login | Password guessing | No | No | No |
If a quiz option describes phishing, password cracking, malware installation, or unauthorized badge cloning, the answer is that option. Those are real attacks. They are simply not tailgating.
Practical controls and the resolution of thresholds
The most effective controls combine hardware, procedures, and culture. Turnstiles and optical mantraps reduce permitted simultaneous entrants to one, physically. A common specification is that the barrier closes fast, often within 1 to 3 seconds, and forces a single-file posture. Antipassback logic in access-control software flags or denies a second badge use at the same portal within a defined interval, typically a few minutes, to catch credential sharing.
Procedures matter more than many products. Require badge display at all times. Ban door holding for unbadged visitors. Route all visitors through a staffed reception desk with escorting. Publish a challenge policy: every employee is expected to question an unbadged person politely, and no one is disciplined for asking.
Detection controls include door-position sensors that alarm if a door stays open beyond its unlock delay, which is usually configurable but often set near 15 to 30 seconds. Video analytics can count entrants per portal and flag mismatches between badge reads and body counts.
The threshold reconciliation matters here. A door held open for 20 seconds is not automatically an alarm; many facilities set the alarm threshold at 30 seconds or longer to accommodate moving carts and deliveries. Antipassback windows of several minutes catch badge reuse without flagging legitimate re-entry after lunch. The short interval (seconds) governs the door hardware; the longer interval (minutes) governs the software logic; the two operate on different layers and should not be conflated.
Finally, the culture control has no threshold and no price tag. It has the highest reported effectiveness against physical tailgating in assessment literature. Hardware buys an alarm. Norms buy compliance.
Limitations of the evidence
Physical tailgating is hard to quantify because it leaves no clean digital trace and few organizations report it publicly. Red-team success rates are real observations but come from selected sites with motivation to test. The label "tailgating" itself varies across standards, vendor documentation, and quizzes, so any incident count depends on the definitions used. Network tailgating variants and physical following share an abstract mechanism but not a common threat model, so controls from one domain transfer imperfectly to the other.
This article is general educational information about security attack categories, not a security certification study guide or a site-specific risk assessment. Do not treat it as a substitute for your organization's access-control policy or for review by a qualified security professional who knows your facility.