\n<\/p>\n
Meta could be the latest company to test the social potential of AI chatbots. <\/p>\n<\/div>\n
A screenshot shared by leaker Alessandro Paluzzi<\/a> on Twitter shows what seems to be an intro screen for the new Instagram feature. It says the chatbots will be able to answer questions, give advice, and help users write messages. It also says that users will be able to choose between \u201c30 AI personalities and find which one you like best.\u201d<\/p>\n<\/div>\n Meta hasn\u2019t announced any formal plans for such a feature, but chatbots would fit past statements about the company\u2019s AI ambitions. In February<\/a>, CEO Mark Zuckerberg said Meta was \u201cdeveloping AI personas that can help people in a variety of ways\u201d and that the company was exploring how to make such bots accessible through text conversations \u201clike chat in WhatsApp and Messenger.\u201d <\/p>\n<\/div>\n Other companies have also seen the potential of chatbots as an engaging social feature. Snapchat launched its own \u201cMy AI\u201d chatbot (powered by OpenAI\u2019s ChatGPT) in February<\/a>. And sites like Character.ai have achieved significant popularity<\/a> by allowing users to train and talk to chatbots based on popular fictional characters. Even bots not intended for social engagement have been adopted for this purpose. When Microsoft launched its Bing chatbot earlier this year, many users were surprised, disturbed, and delighted<\/a> in equal measure by the bot\u2019s strange conversational patter. <\/p>\n<\/div>\n The difficulty for companies, though, is creating a bot that\u2019s engaging and fun to talk to but doesn\u2019t cross the line into offensive or dangerous interactions. For example, not long after Snap released its \u201cMy AI\u201d bot, it began offering disturbing advice<\/a> to users, including encouraging an individual pretending to be 13-years-old to have sex with their 31-year-old \u201cboyfriend.\u201d In March, a Belgian man died by suicide, with his widow claiming that he had been encouraged to kill himself<\/a> by a chatbot he\u2019d been talking to regularly. <\/p>\n<\/div>\n It\u2019s not clear if Meta genuinely intends to launch such bots on Instagram, or what safety steps it may take. We\u2019ve reached out to the company for comment and will update this story if we hear back. The leaker of the above screenshot, Paluzzi, has a reliable track record for spotting upcoming app features, including a BeReal Instagram clone<\/a> and a co-author feature on Twitter<\/a>.<\/p>\n<\/div>\n