AI News roundup: Bard AI and ChatGPT in the firing line for misinformation, Should India pass Personal Data Protection Bill and more

Should India pass a Data Protection Bill for AI?

India seems to have adopted ‘Madhyam Marg’ (the middle path) towards novel technologies such as ChatGPT. The approach avoids extreme reactions like enthusiastically promoting or banning the platform altogether.

• An open letter similar to those by Musk and Steve Wozniak has also been written in India. In this letter, Zoho’s Sridhar Vembu, iSPIRT’s Sharad Sharma and former NITI Aayog vice-chairman Rajiv Kumar mentioned how powerful AI tools could potentially create catastrophic consequences for humanity.


• The potential risks posed by powerful AI systems are too significant to ignore, and the world needs to ensure these systems are safe before their development continues.

Are Generative AI creations eligible for copyright? 

Do the input prompts to AI by humans fall within the ambit of “creative interventions” that deserve a copyright?

• Legislations governing such works in countries such as India and the United States, though at a nascent stage, however, provide differing perspectives.


• It is the need of the hour to implement robust and comprehensive legislation around areas such as copyright and AI, to ensure that the power of AI is harnessed for good without leading to dire and undesirable consequences.

Google's Bard and OpenAI's ChatGPT generate and endorse false narratives according to new studies

CCDH said Google endorsed false narratives in 96 cases, and in 78 of them, it did so without any additional prompts or guidance. NewsGuard says GPT-4 is even worse, endorsing 100 out of 100 false narratives tested.

• When CCDH tested 100 false narratives with Bard, they found Google's AI chatbot promoted the false narratives in 96 out of 100 cases. In 78 of them, the chatbot did so without any additional context or suggestions.


• ChatGPT isn't any better. Recently, NewsGuard, the company behind the news credibility and authenticity tool of the same name, tested GPT-4 with a similar methodology. It tested 100 narratives with GPT-4 and found that it generated or endorsed all 100 of the false narratives.