OpenAI tells ChatGPT models to stop talking about goblins

OpenAI tells ChatGPT models to stop talking about goblinsJust nowShareSaveAdd as preferred on GoogleLiv McMahonTechnology reporterGetty ImagesThe ChatGPT developer discovered the "strange affinity for goblins" while testing tools powered by newer systems, such as its coding agent CodexChatGPT-maker OpenAI has had to instruct some of its AI tools to stop talking about "goblins", after finding the term had randomly crept into responses.In a blog post on Thursday, the company said it spotted increased mentions of the mythological creatures, as well as "gremlins", in ChatGPT, powered by its latest flagship model, GPT-5.After the issue was flagged by users and employees, OpenAI took steps to mitigate it, including telling its coding tool Codex not to refer to goblins unless relevant.However, it highlights challenges AI firms face in tackling the potential for systems and their training to reward and reinforce errors like language quirks.OpenAI said it first noticed increased mentions of goblins, gremlins and other creatures after the launch of GPT-5.1 in November."Users complained about the model being oddly overfamiliar in conversation, which prompted an investigation into specific verbal tics," the company wrote in its blog post on Thursday.It added that after a researcher who had seen a few "goblin" mentions asked it to be checked out, developers found the term's appearance in ChatGPT responses had risen by 175% since GPT-5.1's launch.They meanwhile found that mentions of "gremlin" rose by 52%.The increases, while large, may account for a small amount of responses overall.According to OpenAI, "a single 'little goblin' in an answer could be harmless, even charming," but the uptick in their appearance across output warranted investigation.'Raccoons, trolls, ogres, pigeons'Getty ImagesGoblins weren't the only creature popping up where they shouldn't, OpenAI saidA...