Context: Content on social media and other digital platforms can come under the scanner as the State government will deploy AI-powered Social Media Analytics Solution software to identify hate speech and malicious campaigns in news and advertising.
- Content on social media and other digital platforms could come under the scanner as the Karnataka government will deploy AI-powered ‘Social Media Analytics Solution’ software to identify hate speech, misinformation and malicious campaigns in news and advertising.
- The AI-based software will be used to assess misinformation, disinformation, hate speech and harmful narratives, especially for safety of women, children, and the public.
- The decision to deploy the AI-powered software at the cost of ₹67.26 crore was arrived at during the Cabinet meeting chaired by Chief Minister .
- “This is not only for fact checking, but also to identify manipulation on social media platforms and digital media platforms. Right now, the AI software will be deployed to assess. Law will come later and action will be initiated,”
- On the legal standing of screening of content, there is no absolute bar to screen or verify criminal intentions or manipulation.” He also clarified that this would not be applicable to news media outlets. “Fake media banners and houses will, however, not be spared, that there will be no human interference in this, and that no party or government will benefit.
- The Cabinet note said that the software would be developed adhering to guidelines on data sovereignty and data localisation, and cloud architecture would be developed adhering to the guidelines of MeitY. It also pointed out that the Department of IT/BT has after evaluation found the importance of IDTU project in controlling the spread of misinformation ahead of Lok Sabha elections.
- The proposed software will be proprietary algorithm, which will be dynamic and continuously renewed AI/ ML units since there is a threat of misinformation evolving and fixed algorithm was not an answer. There will be real-time alerts and geo-specific threat mapping. It would help in identifying new-age misinformation, hate speech, deep fakes, bot networks, multi-language narratives, and subject-based manipulations.
Source: The Hindu