Synopsis-
“In order to protect customer privacy, Meta Platforms only used public Facebook and Instagram posts to train its new Meta AI virtual assistant. Private postings shared only with close relatives and friends were not included.”
Data privacy of consumers has become significant in the age of technology, and thus it works as a vital instrument to win the hearts of the target audience. A company like Meta which includes Facebook, Instagram and others mindfully cash these opportunities.
AI Assistant for Meta
In order to protect customer privacy, Meta Platforms only used public Facebook and Instagram posts to train its new Meta AI virtual assistant. Private postings shared only with close relatives and friends were not included.
Nick Clegg, President of Global Affairs, Meta, said, Meta also did not use private chats on its messaging services as training data for the model and took steps to filter private details from public datasets used for training. We’ve tried to exclude datasets that have a heavy preponderance of personal information, adding the “vast majority” of the data used by Meta for training was publicly available”.
The tech company cited privacy concerns as the reason it chose not to use LinkedIn’s content.
Numerous tech firms have come under fire for utilising illegally obtained data to train their AI models, including Meta, OpenAI, and Google’s Alphabet. To summarise data and generate graphics, these models ingest a lot of data.
While dealing with litigation from authors accusing them of copyright infringement, the corporations are pondering how to handle private or copyrighted materials that their AI systems may duplicate.
The first consumer-facing AI tools from Meta were introduced by CEO Mark Zuckerberg at Connect, the company’s annual products conference. Meta AI stood out as the most important of these. This year’s conference, in contrast to previous ones, was predominantly focused on talks about artificial intelligence.
