Why in the news?

Generative AI’s transformative potential challenges existing legal frameworks and judicial precedents, which are inadequate for effectively governing this rapidly-evolving technology designed for a pre-AI world.

What is Generative AI?

• Generative AI refers to artificial intelligence models that can generate original content like text, images, videos, and audio in response to prompts or requests.
• These models learn patterns from large datasets and use that knowledge to create new content that resembles the training data.

Persistent and Contentious Issues in Internet Governance

Safe Harbour and Liability Fixation:

• The Shreya Singhal judgment upholds Section 79 of the IT Act, granting intermediaries ‘safe harbour’ protection against liability for hosted content, contingent upon meeting due diligence requirements.
• Challenges arise in applying these provisions to GAI tools, with debates on whether they should be classified as intermediaries, conduits, or active creators.
• The Delhi High Court’s ruling in the Christian Louboutin Sas vs Nakul Bajaj and Ors (2018) case limited safe harbour protection to “passive” intermediaries.
• The classification of GAI tools complicates the assignment of liability, especially in cases of user reposts.

Generative AI and Legal Conflicts:

• GAI outputs have led to legal conflicts, such as a lawsuit in the U.S. against OpenAI for defamation by ChatGPT.
• Ambiguities in classifying GAI tools complicate legal decisions regarding liability.

What Does the Indian Copyright Act 1957 Say?

• Section 16 specifies that copyright protection is granted only under the provisions of the Act, with reluctance globally to extend protection to AI-generated works.
• Critical questions include whether existing copyright laws should be revised for AI, the need for co-authorship with humans, and the responsibility for copyright infringement by AI tools.
• The 161st Parliamentary Standing Committee Report highlights the inadequacy of the Copyright Act to facilitate AI authorship and ownership.
• Current Indian law allows copyright owners to take legal action against infringement, but liability for AI-generated content remains unclear.

Steps to Pursue (Way forward)

• Learning by Doing: Implement a sandbox approach, granting temporary immunity from liability to GAI platforms for responsible development and data gathering to inform future regulations.
• Data Rights and Responsibilities: Overhaul the data acquisition process for GAI training, ensuring legal compliance, proper licensing, and compensation for intellectual property used in training models. Potential solutions include revenue-sharing or licensing agreements with data owners.
• Simplify the Licensing: Licensing data for GAI is complex due to the lack of a centralized licensing body for web data. The creation of centralized platforms, similar to stock photo websites, can simplify licensing, streamline access to data, and ensure data integrity against bias and discrimination.
• Government and Judicial Approach to maximize the benefits of GAI: A comprehensive re-evaluation of existing digital jurisprudence is needed, requiring a holistic, government-wide approach and judicious interpretations by constitutional courts. The aim is to maximize the benefits of GAI while safeguarding individual rights and protecting against unwelcome harm.

Mains PYQ:

Q The emergence of the Fourth Industrial Revolution (Digital Revolution) hasinitiated e-Governance as an integral part of government”. Discuss. (UPSC IAS/2020)

Get an IAS/IPS ranker as your 1: 1 personal mentor for UPSC 2024

Attend Now