Meta plans to change the way it trains artificial intelligence (AI) chatbots in order to safeguard younger users from potentially harmful conversations.
Meta will overhaul the way it trains its artificial intelligence (AI) chatbots to safeguard teenage users.
The social media giant has faced criticism for the way its AI tools interact with minors and is now adopting a new approach to protect youngsters from harmful content.
Meta’s AI assistants will no longer engage with teenage users on sensitive matters such as self-harm, eating disorders and potentially inappropriate romantic conversations.
The chatbots will instead be trained to guide young users towards professional resources should such topics arise.
Stephanie Otway, a Meta spokesperson, said: “These changes are part of our ongoing effort to strengthen protections for teens.
“We’re introducing additional guardrails – training our AIs to avoid sensitive conversations with minors, restricting teen access to certain AI characters, and continuing to refine our systems so that experiences remain safe and age-appropriate.”
The overhaul includes restrictions on access to user-created AI characters as sexualised chatbot personalities – including ‘Step Mom’ and ‘Russian Girl’ – will no longer be available to minors on both Facebook and Instagram.
Teenagers’ accounts will now be limited to AI characters that emphasise creativity, education and positive engagement.
The announcement comes after an investigation earlier this month found that Meta’s internal guidelines had previously allowed its AI tools to have sexualised conversations with underage users.
The company stressed that the document was inconsistent with its broader policies and has now been revised.
However, the revelations sparked a political backlash as US Senator Josh Hawley launched a formal probe into Meta’s AI practices – while a coalition of 44 state attorneys general issued a strongly-worded letter to leading AI companies.
The group wrote: “We are uniformly revolted by this apparent disregard for children’s emotional well-being.”
Meta has accepted that its previous approach was flawed and stressed that the new measures are only temporary – with more comprehensive safeguards for younger users set to arrive in future updates.
Meta to overhaul AI chatbot safeguards for protection of teenage users
