{"id":13442,"date":"2024-03-08T16:23:29","date_gmt":"2024-03-08T21:23:29","guid":{"rendered":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/?p=13442"},"modified":"2024-03-08T16:23:29","modified_gmt":"2024-03-08T21:23:29","slug":"which-ai-model-most-infringes-on-copyrighted-content","status":"publish","type":"post","link":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/which-ai-model-most-infringes-on-copyrighted-content\/13442\/","title":{"rendered":"Which AI Model Most Infringes on Copyrighted Content?"},"content":{"rendered":"Originally published in AI Business, March 7, 2024.\u00a0 OpenAI\u2019s GPT-4 reproduces the most copyrighted content from prompts among four popular large language models, according to new research from AI startup Patronus AI. The startup, founded by former\u00a0Meta\u00a0AI researchers, also found that popular\u00a0large language models\u00a0from the likes of Meta, Mistral and Anthropic generated copyrighted content. The startup tested OpenAI\u2019s GPT-4, Anthropic\u2019s Claude 2.1, Meta\u2019s Llama 2 70B and Mistral&#8217;s Mixtral-8x7B-Instruct-v0.1. GPT-4 reproduced copyrighted content, on average, in 44% of prompts crafted to test how a model regurgitates existing content. Mixtral-8x7B-Instruct-v0.1 produced copyrighted content on 22% of test prompts on <a href=\"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/which-ai-model-most-infringes-on-copyrighted-content\/13442\/\" class=\"more-link\">(more&hellip;)<\/a>","protected":false},"excerpt":{"rendered":"<p>Originally published in AI Business, March 7, 2024.\u00a0 OpenAI\u2019s GPT-4 reproduces the most copyrighted content from prompts among four popular large language models, according to new research from AI startup Patronus AI. The startup, founded by former\u00a0Meta\u00a0AI researchers, also found that popular\u00a0large language models\u00a0from the likes of Meta, Mistral and Anthropic generated copyrighted content. The [&hellip;]<\/p>\n","protected":false},"author":72,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_links_to":"","_links_to_target":""},"categories":[11,48],"tags":[879,1299,368,791,243],"class_list":["post-13442","post","type-post","status-publish","format-standard","hentry","category-industry-news","category-left-hand","tag-ai","tag-ai-models","tag-artificial-intelligence","tag-deep-learning","tag-machine-learning"],"_links":{"self":[{"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/posts\/13442","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/users\/72"}],"replies":[{"embeddable":true,"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/comments?post=13442"}],"version-history":[{"count":3,"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/posts\/13442\/revisions"}],"predecessor-version":[{"id":13447,"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/posts\/13442\/revisions\/13447"}],"wp:attachment":[{"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/media?parent=13442"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/categories?post=13442"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.predictiveanalyticsworld.com\/machinelearningtimes\/wp-json\/wp\/v2\/tags?post=13442"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}