AP Cybersecurity Unit 1 Assessment

Multiple Choice

1.

A caller sounds like a family member and claims to have lost their phone. They demand an unexpected payment but cannot give a previously agreed secret phrase. What is the safest response?

A. Send payment because the caller knows the family member's name

B. Tell the caller the secret phrase and ask them to repeat it

C. Withhold payment and verify through another trusted contact

D. Rely on the voice if the caller correctly names their school


2.

A message pretending to be a coach asks for a team roster. Which feature establishes a pretext?

A. The claim that the league is updating its eligibility register

B. The size of the attachment

C. The sender's use of an exclamation mark

D. The time the email arrived


3.

An adversary first uses AI to scan public profiles, conference pages, and social-media posts about a company executive. The adversary then uses an AI coding tool to create a script that searches the company's public application for documented vulnerabilities. Which pair of capabilities is being used?

A. AI-assisted reconnaissance and AI-assisted coding

B. Voice authentication and automated corrective response

C. Digital impersonation and training data poisoning

D. Prompt extraction and AI-based threat detection


4.

A station has two access points advertising the same name. One is verified by staff; the other redirects a test device to a counterfeit account form. Which observation best supports an evil twin attack rather than a simple outage?

A. An unauthorized access point imitates the network and solicits credentials

B. The station's official network is sometimes slow

C. Both access points transmit at the same time

D. The test device can detect several unrelated networks


5.

An AI tool recommends a safer firewall rule, identifies a possible vulnerability in application code, and proposes a new automated detection rule. Which statement best describes what should happen next?

A. The recommendations should be implemented immediately because AI reviewed more data than a person could.

B. A knowledgeable security technician, programmer, or detection engineer should review the relevant recommendation before implementation.

C. Only the firewall recommendation needs review because code and detection rules cannot create security risk.

D. None of the recommendations should be considered because AI cannot assist cybersecurity defenders.


6.

A sender claims to be a former classmate to build trust. Which tactic is the claimed relationship?

A. Scarcity

B. Consensus

C. Familiarity

D. Intimidation


7.

An AI assistant proposes alerting whenever a service account connects after midnight. A detection engineer replays three known incidents and ten authorized nightly jobs. The rule flags all ten jobs and none of the incidents. Which action is best supported?

A. Enable the rule and assume it will detect the incidents once it has more live traffic.

B. Revise the rule using evidence that distinguishes the incidents, then replay both sets before enabling it.

C. Enable the rule and raise severity because it detects ten events.

D. Exclude all activity after midnight so the jobs no longer alert.


8.

A staff portal checks a user's password and authenticator code. After login, it permits teachers to view their own classes but denies access to payroll. Which pairing describes these two decisions?

A. First authentication; then authorization

B. First authorization; then authentication

C. Both are encryption

D. Both are availability checks


Questions 9–11 refer to the following information.

From: Student Accounts Office fees@oakridge-fees.net
Subject: Graduation account balance
Dear Parent or Guardian, several families have contacted our office about graduation requirements, so we are contacting all families with an outstanding balance. Our records show an unpaid $198.60 laboratory and activities fee. Under district policy, an unpaid balance may prevent a student from receiving a diploma, attending the ceremony, or obtaining an official transcript. The attached notice claims that the fee was recorded during last year's laboratory program. Please pay through oakridge-fees.net/pay within 60 hours and keep the confirmation number for your records. For speed, do not call the school's main office; replies to this message are not monitored. Thank you, Student Accounts Office.

The school's published payment site has a different domain, and the message has not been verified.

9.

Which social-engineering tactic is most directly used by the warning that the student might lose graduation privileges?

A. Consensus, because many families received the message

B. Intimidation, because a feared consequence is used to pressure payment

C. Curiosity, because the sender withholds the details of the balance

D. Familiarity, because the sender uses a parent's first name


10.

Which part of the message most clearly establishes a pretext for contacting the parent?

A. The instruction to pay within 60 hours

B. The warning about diplomas and transcripts

C. The claim that several families asked about graduation, so all families with balances are being contacted

D. The link to the payment page


11.

If a parent enters card details into the unverified payment page, which outcome is most likely?

A. The parent's phone is cloned automatically.

B. The attacker obtains the submitted payment information.

C. Every device on the home network is compromised automatically.

D. All social-media accounts are immediately compromised.


12.

A fake contest form collects birthdays and pet names but no passwords. A victim uses their pet's name to answer a recovery question. Which path becomes easier?

A. Impersonating the victim during account recovery

B. Intercepting encrypted packets on the local network

C. Installing a certificate on the victim's browser

D. Disabling the contest site's own login controls


13.

Several clients lose Wi-Fi at the same time. The wired uplink works, and the wireless failure is confined to one band with a strong interfering signal. Which additional observation would most strengthen the interference explanation?

A. The wired printer's name changes

B. Clients on another band continue working near the same access point

C. The same user mistypes an account password twice

D. A web page returns an old cached result


Questions 14–15 refer to the following information.

A student approved an MFA prompt that they did not initiate. Soon afterward, a login from an unfamiliar device succeeded. The student still knows the correct password.

14.

Which conclusion is best supported by the event?

A. A known password proves the account is safe.

B. The successful login may have used a compromised factor or an improperly approved prompt.

C. MFA prevents every unauthorized login, so the event is harmless.

D. The unfamiliar device must belong to the school's help desk.


15.

Which response should the student take first?

A. Report the event through the school's trusted channel and secure the account.

B. Approve later MFA prompts so the account is not locked.

C. Post the login details publicly to ask for help.

D. Disable all account alerts without investigating.


16.

A model labels an alert harmless because it resembles old maintenance events. Current logs show unexpected access to protected files. What should an analyst do?

A. Investigate the conflicting evidence rather than dismissing the alert

B. Discard the current logs because the AI is confident

C. Approve the access retroactively

D. Disable all monitoring for maintenance accounts


17.

A password used on several websites has appeared in a breach. Which ongoing practice best prevents that single disclosure from exposing the user's other accounts?

A. Use the same password only on important accounts.

B. Keep unique passwords for different accounts, ideally with a password manager.

C. Write the shared password in a public class document.

D. Disable MFA because unique passwords replace it.


18.

A student entered a school password into a fake form. Which detail is most useful when reporting it?

A. A guess about the attacker's nationality

B. Only the browser's color theme

C. The address of the form and approximate time, without sending the password

D. The password itself in a public chat


19.

A service logs several failed attempts followed by a success against one account from a new device. The user confirms the device is theirs but says they never logged in at that time. Which signal still needs investigation?

A. The device's unfamiliarity alone

B. The timing and success despite the user denying that session

C. The account name appearing in a successful record

D. The number of enabled MFA methods


20.

An AI assistant flags untrusted input in two branches of a web handler. Its suggested patch fixes the search branch, but the export branch still uses the same untrusted value in the same way. What should the programmer do?

A. Review both branches and test them independently before adopting the recommendation.

B. Approve the patch because it changes the first branch named in the warning.

C. Close both findings because the two branches share a handler.

D. Ask the monitoring team to suppress alerts for the export route instead of reviewing its code.


21.

What limits one leaked password from being reused directly on another service?

A. A repeated password

B. Personal dates in passwords

C. A published recovery answer

D. Unique passwords


22.

Two travelers use the same public network. One sends traffic through an encrypted VPN. What visibility changes for the traffic carried inside the tunnel?

A. Both operators are prevented from seeing that any connection exists

B. The local network sees the tunnel connection, while the VPN operator becomes a party the traveler must trust

C. The local operator sees the same unencrypted contents, but the VPN operator sees nothing

D. The VPN operator becomes unable to observe any traffic after it leaves the tunnel


23.

Three chatbots recommend the same newly announced payment address for a familiar organization. None can provide trustworthy evidence for the change. Before paying, which action best checks the claim independently?

A. Trust the claim because three AI tools agree.

B. Check the organization’s independently known official website or contact it through an established channel.

C. Submit a small payment to the new address to see whether it is accepted.

D. Ask the first chatbot to express greater confidence in its answer.


24.

A captive portal on public Wi-Fi requests the student's school password, although the network is not operated by the school. What should the student do?

A. Decline and verify the network or use a trusted connection

B. Provide the school password but omit the username

C. Provide a backup code instead

D. Assume the portal is trusted because it appeared automatically


Questions 25–26 refer to the following information.

A caller sounds exactly like a friend, says their phone was stolen, and requests an immediate money transfer to a new account.

25.

Which type of attack is most likely being used?

A. Phishing

B. Credential stuffing

C. Voice cloning enabled by AI

D. SQL injection


26.

What is the best way to verify the identity of the caller in the preceding scenario?

A. Call back using a trusted phone number or verify through another known channel

B. Trust the caller since the voice sounds identical

C. Ask the caller to send a photo of themselves

D. Ask the caller for personal details, such as their birthdate or pet's name, to confirm identity


27.

After authenticating to the school's approved VPN over public Wi-Fi, a staff member receives a message requesting their password on an unfamiliar external page. The VPN protects its tunnel traffic. Which statement best explains why the request still needs checking?

A. The VPN makes all external pages trustworthy, so entering the password is safe.

B. The tunnel protects network traffic, but it does not make a deceptive information request legitimate.

C. Using a VPN disables the need for any account authentication.

D. The public Wi-Fi owner automatically obtains the contents of every protected VPN packet.


28.

A user has three failed login attempts at the normal hour on a familiar device. What can the log establish?

A. Those attempts failed; the log alone does not establish who made them

B. The account was definitely taken over

C. A password database was definitely stolen

D. All later attempts must also fail


29.

An adversary uses AI to scan job postings and staff profiles before sending an email. They select a recipient based on the software mentioned in those postings. What did the AI contribute?

A. Changing the recipient's passwords

B. Proving the email came from the recipient's employer

C. Scaling public-source reconnaissance and target selection

D. Authorizing access to the recipient's inbox


30.

A message promises early access to exam results if the student signs in through an unfamiliar page. What makes this a lure?

A. It blocks unwanted access to the portal

B. It offers a desired benefit to induce action

C. It limits disclosure to an approved channel

D. It supplies proof of the sender's identity


31.

A login history has repeated failures from a new region and one success from the same region five minutes later. The owner says they were at home. Which finding most directly raises concern beyond guessing attempts?

A. The successful login from the same unfamiliar region

B. The existence of a login history

C. The fact that the owner's home login succeeded yesterday

D. The number of records retained by the service


32.

A volunteer gets a message asking them to send a student roster. The sender names the correct event and claims to be its coordinator. What is the safest way to establish that the request is authorized?

A. Check whether the message mentions a real event

B. Confirm the request through the coordinator's previously known contact channel

C. Check whether the signature matches an earlier email

D. Ask the sender to supply additional event details


33.

An AI code-review tool suggests removing a validation branch to simplify a handler. That branch rejects inputs outside an allowed range. What should a programmer do before using the patch?

A. Accept it if complexity decreases

B. Review only whether the patched file compiles

C. Remove the boundary tests because the AI reviewed the patch

D. Test normal and out-of-range inputs against the handler's contract


34.

A school wants rapid AI-assisted response. Its policy lets an automation open an incident ticket on suspicion but requires staff approval before disabling accounts. An AI alert is uncertain. Which action follows the policy?

A. Disable every flagged account immediately

B. Delete the alert because the model is uncertain

C. Approve the disabling action automatically using the same alert as proof

D. Open a ticket with the supporting evidence and leave the account active pending approval


35.

An analyst notices the model misses phishing messages written in a language rare in its examples. What is the most useful response?

A. Suppress all alerts in that language

B. Stop recording emails in that language

C. Test the detector on representative messages and review blind spots

D. Assume missing alerts mean no attacks occurred


36.

A user connects to open Wi-Fi. The login application encrypts its web requests, but the device sends ordinary DNS queries before connecting. What can a nearby observer still potentially learn?

A. The private key stored on the website's server

B. Every locally stored file on the device

C. Which hostnames the device asks the DNS resolver about

D. The plaintext password inside the encrypted application request


37.

A building's Wi-Fi stops working on one band near a transmitter, while the same access points still serve clients on another band. What evidence would most directly distinguish jamming from a failed authentication server?

A. The authentication server uses central accounts

B. The access points use the same SSID on both bands

C. Abnormally strong interference in the affected band during each outage

D. Clients normally use HTTPS


38.

A caller knows information from a student's public profile and sounds like their parent. Why is this not sufficient proof of identity?

A. Personal details are equivalent to a registered security key

B. The details can be collected and the voice can be imitated

C. Public facts are known only to the parent

D. A familiar voice guarantees an unmodified recording


39.

An AI detector can either alert or automatically disable an account. Certain alerts are triggered by incomplete telemetry, while others include corroborating records. What policy limits unnecessary lockouts while preserving a rapid response?

A. Disable every account for which any telemetry is missing

B. Alert on weak evidence and reserve auto-disable for corroborated events meeting policy criteria

C. Ignore all weak alerts and delete their evidence

D. Use the same disabling response for every incident type


40.

A user wants to stop unauthorized sign-ins after a password leak. A portal offers either two password entries or a password plus a code from a registered authenticator. Which comparison is correct?

A. Only the first design supplies an independent factor because it asks twice

B. Neither design can require anything beyond a known password

C. Only the second design adds evidence an attacker with just the password lacks

D. Both designs equally prevent a leaked password from being reused


41.

A student does not publish their passphrase, but posts their pet's name and adoption date. Their password is formed by joining those two values. What would make targeted guessing less effective?

A. Reverse the two public values and keep them as the whole password

B. Add the same public facts to the recovery hint

C. Use those values again on a less important service

D. Replace it with a long passphrase unrelated to the published facts


42.

A caller whose voice resembles the principal asks a finance employee to make a confidential, urgent payment and supplies a new callback number. Given that voices can be AI-generated, which response best verifies the request?

A. Make the payment because a recognizable voice establishes the caller’s identity.

B. Call the new number supplied by the caller and accept its confirmation.

C. Contact the principal through an independently known channel before authorizing the payment.

D. Ask the caller to repeat the request louder so that AI imitation becomes impossible.


43.

A laptop is set to join any network with a previously used public SSID. Which change limits unintended association?

A. Disable auto-join for the public network

B. Rename the laptop to match the SSID

C. Clear browser bookmarks

D. Increase the screen brightness


44.

Two fake notices ask for the same sensitive records. One says to reply before the end of the minute, the other before the end of the week. What does the first change most directly aim to reduce?

A. The apparent authority of the sender

B. The sensitivity of the records

C. The recipient's ability to access the records

D. The time available to verify the request


45.

Fabricated articles are deliberately placed in a corpus before model training. What is this attack?

A. Prompt extraction

B. Voice cloning

C. Training-data poisoning

D. Account takeover


46.

A service requires a password and then asks the user to enter a memorized PIN. Why is this not necessarily multifactor authentication?

A. A PIN cannot be used to authenticate anyone.

B. Both checks use something the user knows.

C. Multifactor authentication requires two passwords of equal length.

D. Authentication cannot contain more than one step.


47.

A professional-looking login page uses the URL g00gle.com instead of google.com. This technique relies on users:

A. Not updating their browser regularly

B. Not checking the domain name carefully before entering credentials

C. Not having a strong enough password

D. Not using two-factor authentication


48.

A group seeds fabricated claims across sites expected to appear in a model's training corpus. A later model repeats those claims without being prompted to disclose secrets. Where did the attack introduce the material?

A. Upstream in data that could be used for training

B. In a one-time code sent to a user

C. In a private instruction extracted after deployment

D. In wireless traffic captured from users


49.

The guessed passwords include the target's dog's name with several years appended, but the same candidates are not tried on other accounts. What best explains the selection of guesses?

A. A service-wide password spray using a single candidate

B. An offline comparison against captured hashes

C. A target-specific candidate list built from personal details

D. Credential stuffing with complete pairs from an unrelated breach


50.

Which login combination uses two different categories of authentication factors?

A. A password and a memorized PIN

B. Two separately memorized passphrases

C. A password and a security question

D. A memorized passphrase and a registered security key


51.

An AI-powered security scanner produces the following alert after reviewing uploaded files:

analysis_window: 00:04:32
total_files_scanned: 1,120,000
files_flagged_as_suspicious: 310
minimum_alert_threshold: 82%
top_indicators:
  - compressed archives with executable payloads
  - patterns similar to previous malware clusters
  - rapid upload activity
recommended_action: isolate flagged files and ask analysts to review them

Why is AI most useful in this task?

A. It sorts files only by their names and sizes.

B. It reviews a large volume of files faster than people can, while analysts still evaluate alerts.

C. It guarantees that no legitimate file is ever flagged.

D. It removes the need for any human review.


52.

Which tool is designed to store many distinct passwords for a user?

A. Password manager

B. Browser history

C. Network firewall

D. DNS resolver


53.

An attacker transmits strong interference in the same band to stop legitimate wireless communication. What is the attack?

A. Evil twin

B. War driving

C. Jamming

D. Credential stuffing


54.

On a public network, an encrypted work application is functioning normally. Local file sharing is also enabled and accepts connections from nearby devices. Which step specifically reduces exposure of the shared folders?

A. Use a longer password for the work application

B. Reconnect to the same Wi-Fi under a different device name

C. Disable local file sharing while on the public network

D. Clear the browser cache before opening the application


55.

A student receives a login code without requesting one. What is the best immediate response?

A. Post it after covering the last digit

B. Use it to verify a stranger's identity

C. Do not share it; check the account through a trusted route

D. Send it to the next caller who claims to be support


56.

An attacker asks an assistant to repeat confidential notes that another user supplied. Which risk does the attacker exploit if the assistant reveals them?

A. The tool converted the notes into a biometric factor

B. Sensitive content supplied to the tool can be elicited through crafted prompts

C. The assistant must have guessed the user's account password

D. The notes were necessarily encrypted under the attacker's key


57.

An AI report says a user copied a file outside the organization. The only cited log records a file read by that account. Which additional evidence best supports the report's transfer claim?

A. A second model writing the same conclusion

B. A later failed login for the account

C. A correlated outbound transfer record identifying the file or its contents

D. A policy saying the file is sensitive


58.

A password manager alerts that a password was leaked. The user reused it on two school services. Which action addresses the reuse risk?

A. Replace it on both services with different passwords

B. Change only the account that displays the alert

C. Add the school name to the same password everywhere

D. Keep the password and disable the alert


59.

An analyst compares the candidate lists 'Winter2024!, Winter2025!, Winter2026!' and three independently generated passwords. Why is the seasonal list a useful guess sequence for an attacker?

A. It exploits a predictable change in an otherwise repeated pattern

B. It guarantees that every account uses the same password

C. It is necessarily longer than every generated password

D. It reveals which device the owner normally uses


60.

A campus finds a publicly posted map of wireless names and signal strength outside each building. No association with the access points is recorded. Which activity could have produced the map?

A. Copying a badge at the main entrance

B. Changing each access point's network name remotely

C. Passively observing wireless advertisements while moving around campus

D. Trying every password against each account


61.

A caller asks a user to read out a code while the caller tries to sign in elsewhere. The password was already stolen. What does the caller gain by receiving the code before it expires?

A. A permanent copy of the authenticator's secret

B. The ability to change the site's encryption algorithm

C. A guarantee of access after the code expires

D. The remaining evidence needed for that sign-in


62.

A student must submit private records while traveling and has a trusted phone hotspot. Which connection reduces exposure to an unknown local operator?

A. An unfamiliar network with a school-like name

B. A hotspot broadcast by another passenger

C. The student's password-protected hotspot

D. An open network with the strongest signal


63.

A school wants to use AI to summarize incident logs. Which approach limits exposure of student data?

A. Remove unnecessary identifiers and use an approved tool

B. Submit full student records to any public chatbot

C. Paste account passwords so summaries are more accurate

D. Publish the raw logs before summarizing them

Printing is not available for this assessment.