A financial institution uses an AI model to approve loan applications. The model is trained on historical data that includes biased decisions. The institution also wants to protect the model from adversarial attacks that could manipulate loan approvals. Which approach best addresses both concerns?

Deploy model inversion prevention and output filtering

Use content provenance and input sanitization

Implement robust model training and bias testing

Apply data poisoning detection and user verification

CiscoAI Technical Practitioner

Domain 3Objective 3

3.3 Explain AI-specific Security Threats and Risks, Including Misinformation AI-TECHNICAL-PRACTITIONER Practice Questions (Page 5)

Part of the Ethics and Security domain, which accounts for 15% of the AI-TECHNICAL-PRACTITIONER exam. Cisco does not publish an official question count, but from its 60-minute exam (~25–40 total, ~4–6 in this domain), expect 1–2 from this objective — we provide 22 practice questions to prepare you well beyond it. (estimate)

22questions here

5free pages

4concepts

15%of the exam

Questions 21–22

21expert · hard
A financial institution uses an AI model to approve loan applications. The model is trained on historical data that includes biased decisions. The institution also wants to protect the model from adversarial attacks that could manipulate loan approvals. Which approach best addresses both concerns?
Select an answer first
22foundation · easy
An AI system generates a realistic news article containing false information that is then shared widely on social media. What is the primary risk associated with this scenario?
Select an answer first

Finished these 2 questions?

Review the revealed explanations, or continue through the curriculum.

No more pagesBack to AI-TECHNICAL-PRACTITIONER

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Cisco. “AI-TECHNICAL-PRACTITIONER” is a trademark of its owner, used for identification only.

3.3 Explain AI-specific Security Threats and Risks, Including Misinformation AI-TECHNICAL-PRACTITIONER Practice Questions (Page 5)

Questions 21–22

A financial institution uses an AI model to approve loan applications. The model is trained on historical data that includes biased decisions. The institution also wants to protect the model from adversarial attacks that could manipulate loan approvals. Which approach best addresses both concerns?

An AI system generates a realistic news article containing false information that is then shared widely on social media. What is the primary risk associated with this scenario?

Keep studying — Ethics and Security