Which of the following is the BEST control for preventing deepfakes?

The correct answer is A. Output provenance verification. Output provenance verification (A) is the most direct and effective technical control against deepfakes. Provenance verification uses cryptographic signing, watermarking, or metadata attestation to confirm the authentic origin of media content - allowing recipients to verify…

AI Security Design and Implementation

Question

Options

AOutput provenance verification
BRegular AI risk assessment
CAI governance policies
DSystem input validation

How the community answered

(35 responses)

A
80% (28)
B
3% (1)
C
11% (4)
D
6% (2)

Explanation

Output provenance verification (A) is the most direct and effective technical control against deepfakes. Provenance verification uses cryptographic signing, watermarking, or metadata attestation to confirm the authentic origin of media content - allowing recipients to verify that content was produced by a legitimate source and has not been synthetically generated or altered. Option B (AI risk assessment) is a periodic process, not a real-time prevention control. Option C (governance policies) establishes rules but doesn't technically detect or prevent deepfakes. Option D (input validation) controls what goes into a system, not what comes out - and deepfakes are an output-side concern.

Topics

#Deepfakes#Provenance verification#Synthetic media#AI security controls

Community Discussion

No community discussion yet for this question.

Full AAISM Practice