CDT sees policy takeaways in OpenAI’s ‘post-mortem’ of flawed ChatGPT update

OpenAI’s detailed account of a flawed update to its ChatGPT model deserves praise for transparency but it also raises important questions about how tech companies approach AI safety in the absence of required controls, according to a key digital-rights stakeholder.

“It’s a really interesting post-mortem, it’s interesting that this wasn’t even a major model update, it was a relatively small update,” commented Miranda Bogen of the Center for Democracy and Technology. “It points to the challenge of determining which risks...