A recent report by NewsGuard highlights significant concerns regarding misinformation generated by leading AI tools, OpenAI’s ChatGPT and Google’s Bard. The audit found that these generative AI platforms displayed an alarming 80-98% likelihood of producing false claims on various news topics. In testing, ChatGPT produced 98 out of 100 myths from NewsGuard’s database of false narratives, while Bard generated 80. This raises critical questions about the trustworthiness of AI-generated content. In response, the White House announced large-scale testing of these models at the Def Con 31 conference starting August 10, aiming to enhance the trust and safety framework for generative AI technologies. This initiative will involve thorough evaluations by AI experts and community partners, emphasizing the need for accountability in AI development. As misinformation remains a pressing issue, ensuring reliable AI outputs is essential for media integrity and public trust.
Source link
Share
Read more