The Ministry of Electronics and IT (MeitY) is engaged on a synthetic intelligence (AI)-specific legislation, below which it could not prescribe any penal penalties for violations, in a recognition that the expertise’s advantages outweigh its downsides, The Indian Specific has learnt.
The legislation, which could possibly be launched as a standalone piece of laws, might require social media platforms like Fb, Instagram, YouTube, and X so as to add watermarks and labels to content material that has been generated utilizing AI.
The ministry can be trying on the authorized parameters below which it may possibly ask corporations constructing giant language fashions to coach their databases on Indian languages and “India-specific” content material in order that the fashions can perceive the “Indian context” higher, a senior authorities official stated.
That is being checked out after some AI techniques, like Google’s Gemini, have been discovered to offer totally different solutions for a similar form of queries involving world leaders and Prime Minister Narendra Modi final yr. The IT Ministry didn’t reply to a request for remark.
Final November, Union IT Minister Ashwini Vaishnaw had stated that the federal government will type rules to regulate the unfold of deepfakes on social media platforms, terming them a “new risk to democracy.”
The election cycle earlier this yr stopped work on the legislation, which is now understood to have discovered newfound momentum inside the ministry. On the time, Vaishnaw had stated the Centre’s plan can have 4 key pillars: detection of deepfakes, their prevention by eradicating or decreasing their virality, strengthening reporting mechanisms, and spreading consciousness concerning the expertise.
In Could, IT Secretary S Krishnan throughout an business occasion had stated that whereas the federal government will try to manage AI, it is not going to accomplish that at the price of innovation, signalling the truth that the ministry is predicted to take a center path for AI regulation, because it has taken in a few of its current expertise legal guidelines equivalent to information safety, the place the main target isn’t just on defending civil rights but in addition on facilitating the nation’s rising start-up base.
Commercial
“Whereas we’ll try to manage AI, we’re clear that innovation is just not stifled within the course of. It (innovation) must be inspired… Like we did with the DPDP (Digital Private Knowledge Safety) Act, we’ll make sure that each the pursuits of innovation and safety of important pursuits will are available in sooner or later,” Krishnan added.
To make sure, some social media and AI corporations have stated that they’ve began watermarking AI-generated content material on their platforms earlier than rules require them to take action. Adobe, Arm, Intel, Microsoft, and Truepic have additionally created a joint challenge known as the Coalition for Content material Provenance and Authenticity (C2PA), which focuses on techniques to offer context and historical past for digital media.
Meta has stated that it’s constructing instruments that may establish “invisible markers” in content material items counting on requirements set by the C2PA so it may possibly label photos from Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock.
Adblock check (Why?)
