Spot the Deepfake
Cybersecurity professionals should treat unexpected videos, audio clips, or messages requesting sensitive information with this security principle (check your new CSAM pins!):
"____ AI? Verify!"
What is trust?
This two-letter abbreviation stands for Artificial Intelligence.
What is AI?
DAILY DOUBLE:
AI can help organize your Monday morning, but unfortunately, it still can't provide you with this essential workplace beverage.
What is coffee? ☕😄
True or False: AI can make mistakes.
What is TRUE?
Before entering company information into an AI tool, employees should first determine whether this type of information is permitted to be shared.
What is sensitive or confidential information?
A coworker receives an unexpected message from the CEO requesting sensitive information. They should do this before responding.
What is verify the request through an approved channel?
Also accepted: What is to report to their manager?
Cybercriminals use AI to create highly convincing ________ emails faster than ever.
What is phishing?
TRUE or FALSE: If an AI-generated answer sounds professional, it must be accurate.
What is FALSE?
DAILY DOUBLE:
"Here's my password. Can you save it for later?" Your answer should be this two-letter word.
What is NO? 😄
The instructions or question you give a generative AI tool are commonly called this.
What is a prompt?
This type of AI creates new text, images, audio, or other content.
What is Generative AI?
Putting confidential company information into an unauthorized public AI tool can create this type of risk.
What is a data security/privacy risk?
also accepted: What is a data leak/breach?
An AI system repeatedly producing unfair outcomes for certain groups could indicate this problem.
What is bias?
This AI technique allows computers to learn patterns from data instead of being explicitly programmed for every task.
What is machine learning?
When an AI system generates false or inaccurate information, it is called this.
What is a hallucination?
The principle of giving a user only the system permissions necessary to perform their job is this security principle.
What is the principle of least privilege?
AI-generated answers should undergo this process before being relied upon for an important business decision.
What is human review or validation?
Attackers use AI in this method when hacking to automatically identify and exploit vulnerabilities.
What is AI-assisted hacking?
The CIA triad is an acronym concerning three important concepts in regard to Cybersecurity.
What is Confidentiality, Integrity, & Availability?
A coworker receives an urgent voice message that sounds like an executive asking for an unusual financial transaction. AI could have been used to create this type of synthetic media.
What is a deepfake?
You should never provide any of these in response to an AI Agent. (Provide at least 3)
Correct answers include:
What are users IDs?
What are passwords?
What is sensitive information?
What is PII/PHI?
AI-generated answers should not automatically be trusted simply because they sound confident. Employees should always apply this human skill when reviewing AI-generated answers.
What is critical thinking?
This should be done before placing an AI-generated statistic in a presentation.
What is fact-checking/verification?
Keeping a person involved in reviewing or approving important AI-generated decisions is commonly described by this phrase.
What is human-in-the-loop?