– Meta testing AI chatbot in India across WhatsApp, Instagram, and Messenger
– Company blocking specific election-related queries in chatbot test phase
– Meta working to improve AI response system and restricting certain keywords for accuracy
Last week, Meta began testing its AI chatbot in India across WhatsApp, Instagram, and Messenger. However, with the Indian general elections beginning, the company is already blocking specific queries related to elections in its chatbot. Meta confirmed that it is restricting certain election-related keywords for AI in the test phase and is working to improve the AI response system.
Critics have expressed concerns that generative AI, like Meta’s chatbot, could provide misleading or false information to users, impacting the democratic process. In response to these concerns, Meta is proactively curtailing the scope of its generative AI services and blocking certain election-related queries. Google has also begun blocking election-related queries in its Gemini chatbot experience in India and other markets with upcoming elections.
Meta’s approach to handling GenAI queries involves redirecting users to the Election Commission’s website when specific election-related terms are mentioned. While responses to party names are not strictly blocked, queries containing the names of candidates or certain terms might receive a redirect response.
The company recently rolled out a new Meta AI chatbot powered by Llama 3 in several countries, excluding India. Meta stated that the chatbot will remain in the test phase in India for now, and it is learning from user tests in the country. Currently, Meta AI is not blocking election-related queries for U.S.-related terms, but it remains unclear if restrictions will be implemented during the U.S. elections or in other markets.