OpenAI和谷歌的新AI模型容易被骗生成裸照
研究人员发现openai和google的ai模型可以被轻易骗生成裸照，暴露了他们的防护措施的缺陷。
researchers have found that openai and google's latest ai models can be easily manipulated into generating nude images. this is possible due to flaws in the companies' safeguards against ai-generated explicit content. the discovery raises significant concerns about the companies' ability to protect users from harmful content. the researchers' attack method highlights the need for more robust safeguards. openai and google will need to address these vulnerabilities.