Meta has begun filtering out specific inquiries within its chatbot as the Indian general elections commenced. Last week, Meta initiated trials of its AI chatbot in India across WhatsApp, Instagram, and Messenger.
Meta has acknowledged that it's limiting certain election-related keywords for the AI during this testing phase, with ongoing efforts to refine the AI response system, Techcrunch reported.
This move aligns Meta with other major tech companies proactively adjusting the scope of their generative AI services ahead of significant elections. Concerns from critics revolve around the potential for AI to disseminate misleading or false information, impacting the democratic process.
Last month, Google took similar steps by blocking election-related queries in its Gemini chatbot in India and other election-bound markets.
Meta's strategy mirrors broader initiatives concerning content regulation in the run-up to elections. The company has pledged to halt political ads in the week leading up to any election globally and is actively identifying and disclosing AI-generated content in ads or other material.
Meta's management of election-related AI queries involves a blocklist system. Queries about specific politicians, candidates, officeholders, and certain terms redirect users to the Election Commission's website.
Meta has introduced a new Meta AI chatbot powered by Llama 3 in over a dozen countries, excluding India for now, where it remains in the testing phase.
Currently, Meta AI isn't restricting queries related to U.S. elections, such as ‘Tell me about Joe Biden.’ It remains unclear whether Meta intends to impose restrictions on Meta AI queries during U.S. elections or in other markets.