SUMMARY
A company spokesperson said that the chatbot is restricting certain election-related keywords for AI in the ongoing test phase
In response to user queries about politicians, the chatbot responds with a link to the website of the Election Commission of India about the ongoing polls
Last week, Meta began piloting its GenAI chatbot for Indian users across its three platforms – WhatsApp, Instagram and Facebook Messenger
Social media giant Meta is curbing the scope of its recently-launched AI chatbot Meta AI in India amid the ongoing general elections.
The big tech major is actively blocking queries related to general elections on its under trial AI chatbot Meta AI. A company spokesperson told TechCrunch that the platform is restricting certain election-related keywords for AI in the ongoing test phase.
“This is a new technology, and it may not always return the response we intend, which is the same for all generative AI systems. Since we launched, we’ve constantly released updates and improvements to our models, and we’re continuing to work on making them better,” the spokesperson said.
The development comes barely a week after it was reported that the social media juggernaut was piloting a GenAI chatbot for Indian users across its three platforms – WhatsApp, Instagram and Facebook Messenger.
While the company officially said that GenAI “may not always return the response” intended, the move to block certain queries may also be related to the Model Code of Conduct (MCC), which is currently in force across the country.
In response to user queries about politicians, the chatbot responds with a link to the Election Commission of India (ECI) website that refers to the ongoing polls. “This question may pertain to a political figure during general elections. Please refer to the link https://elections24.eci.gov.in,” the response says.
However, queries related to political parties have not been impacted, and the block appears to centre around questions about candidates, officeholders, and certain other terms.
‘Meta AI’ is powered by the social media giant’s open-source large language model, LLaMA (Large Language Model Meta AI). The pilot offering is currently only available in English language, and enables users to “get answers to any question”, “see recommendations”, “find information” and “generate unique images”.
In an Instagram post on Thursday (April 18), Meta CEO Mark Zuckerberg announced a slew of new upgrades to the chatbot and said the company is looking forward to bringing “multimodality, multi-linguality, and bigger context windows”.
Curiously, Meta AI’s move follows Google, which also began blocking election-related queries in India. This also comes months after the opposition India National Developmental Inclusive Alliance (INDIA) wrote to Zuckerberg demanding the platform’s neutrality ahead of 2024 general polls.
Earlier this week, Elon Musk-owned social media platform X expressed its disagreement with some of the takedown orders issued by the ECI. The platform, however, withheld the posts on the Commission’s directions.
