According to a recent consultation on AI technologies, the government rejected the need to license untested AI models, but insisted on the need to label AI-generated content. Instead of approving the artificial intelligence models under development, the new advice issued by the Ministry of Electronics and Information Technology on Friday evening clarifies the compliance requirements according to the 2021 IT rules.
"This warning has been issued to withdraw the prospectus... Effective March 1, 2024"
Under the new guidelines under the Information Technology (Interim Guidelines and Digital Media Code of Ethics) Rules 2021, IT companies and platforms are often lax in their duty of care.
The government required companies to label content created with their software or AI platforms and inform users of any inherent inaccuracies or unreliability of the results obtained with their AI tools.
“If an intermediary allows or facilitates, through its software or other computer resources, the creation, generation or artificial modification of textual, auditory, visual or audiovisual information and if such information is potentially inaccurate or information generated or modified by software. or other computer resources that may be used as deepfakes must be identified as having been created or modified using the intermediary computer resources. We recommend installing it.”
He added that when changes are made by the user, the metadata must be configured to identify the user or computing resource that affected those changes.
The government on March 1 ordered social media and other platforms to flag AI models in clinical trials as illegal content following a controversy over Google's AI platform's response to Prime Minister Narendra Modi's questions. It was recommended to avoid web hosting.
The Ministry of Electronics and Information Technology warned intermediaries and platforms of criminal prosecution for non-compliance.
Previous recommendations required government approval for the deployment of experimental or untrustworthy artificial intelligence (AI) models and required government approval for the deployment of experimental or untrustworthy artificial intelligence (AI) models. Companies have been asked to adopt the technology only after labeling. as “may be unreliable.”
0 Comments