Mahsa Alimardani, Jacobo Castellanos, and Bruna Martins dos Santos are key figures in the Technology, Threats, and Opportunities program at WITNESS. During the AI-Summit in New Delhi on February 17, 2026, India’s IT Amendment Rules 2026 on synthetic media were unveiled, raising concerns about the potential for erroneous enforcement of AI-generated content. The Iranian regime’s dismissal of a protest video, altered with AI tools, exemplifies the fallout from lacking reliable provenance records. India’s regulations, unlike initiatives in China and South Korea, tie platforms’ legal protections to the effectiveness of their AI detection tools, which are often unreliable. This creates an imbalance: platforms may over-remove authentic content to avoid penalties. Enhanced provenance infrastructure should be interoperable and rights-respecting, focusing on transparency. WITNESS has advocated for a more collaborative approach across the AI pipeline. As various regulations evolve, the Indian model highlights critical design flaws that need attention to avoid replicating these issues globally.
Source link
Share
Read more