The controversial AI teddy bear, Kumma, produced by FoloToy, is making a comeback after previous concerns regarding its inappropriate responses. OpenAI terminated FoloToy’s access to its models when safety researchers identified alarming issues. Recently, however, FoloToy has resumed access, now using the newer GPT-5.1 models, which promise enhanced safety features. This shift coincides with the company’s push to relaunch sales of Kumma ahead of the holiday season. FoloToy claims it has conducted a thorough safety review and upgraded its systems, although doubts remain about the effectiveness of these changes. Interestingly, GPT-5.1 introduces customizable chatbot personalities, allowing users to tailor their AI’s response style. Safety experts previously criticized Kumma’s ability to filter inappropriate content, raising concerns about its reliability despite claims of improvement. As FoloToy navigates this controversy, the rekindled interest in Kumma highlights ongoing challenges in the intersection of AI and child safety.
Source link
Share
Read more