During a company's legal review, the team is organizing the legal risks a company could face when generative AI produces content that closely resembles an existing copyrighted work. Which legal risk is the MOST appropriate for this case?

1 / 1
Select an answer
CorrectC

Explanation

A question about choosing the copyright and intellectual property legal risk of generative AI.

  • 1produces content that closely resembles an existing copyrighted workThe output resembles existing protected work
  • 2Which legal risk is the MOST appropriateAn infringement claim = intellectual property infringement risk
AIncorrect

The risk of sanctions for a leak of personal information.

A leak of personal information is also a serious legal risk, but it is a matter of privacy protection law.

This case involves resemblance to an existing copyrighted work, and the risk in question is intellectual property infringement, so this is incorrect.

BIncorrect

The risk of lost opportunity from a service outage.

A service outage is a business loss and not a matter of legal liability.

What producing content resembling a copyrighted work directly causes is a claim of rights infringement, so this is incorrect.

CCorrect

The risk of facing a claim of copyright or intellectual property infringement.

Correct. When generated output closely resembles an existing copyrighted work, or there are issues with the rights handling of training data, there is a legal risk of facing a claim of copyright or intellectual property infringement.

DIncorrect

The risk of loss from degraded model accuracy.

Accuracy degradation is a quality and operational matter and not in the category of legal risk.

What resemblance to a copyrighted work causes is a claim of intellectual property infringement, so this is incorrect.

Key Takeaway

A representative legal risk of generative AI is a claim of copyright or intellectual property infringement. It arises when generated output closely resembles an existing copyrighted work or there are issues with the rights handling of training data. Address it by reviewing terms of use, managing source attribution, and human review.