@misc{somepalli_diffusion_2022,
	title = {Diffusion {Art} or {Digital} {Forgery}? {Investigating} {Data} {Replication} in {Diffusion} {Models}},
	shorttitle = {Diffusion {Art} or {Digital} {Forgery}?},
	url = {http://arxiv.org/abs/2212.03860},
	doi = {10.48550/arXiv.2212.03860},
	abstract = {In this paper, Somepalli et. al. investigate the extent of digital forgery in various cutting-edge diffusion AI models, whose high quality and customizability enables them to be used for commercial art and graphic design purposes. Their findings suggest that factors such as training set size impact the rate of content replication, and that in the case of the popular Stable Diffusion model, the training data is blatantly copied to produce output.

As AI art generator tools such as DALL.E and Stable Diffusion are being increasingly used for commercial purposes, the potential for AI to replicate copyrighted content without proper attribution could have significant legal implications for lawmakers, particularly as it relates to the copy-right and intellectual property regulation of AI art generators, necessitating a reevaluation of current copyright laws in the context of AI-generated art and design. (Methods and Metrics, Detecting digital (art) forgery)},
	urldate = {2024-01-26},
	publisher = {arXiv},
	author = {Somepalli, Gowthami and Singla, Vasu and Goldblum, Micah and Geiping, Jonas and Goldstein, Tom},
	month = dec,
	year = {2022},
	note = {arXiv:2212.03860 [cs]},
	keywords = {20, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Computers and Society, Computer Science - Machine Learning, Trustworthy AI, gw\_abstracts},
}

{"_id":"ivEqtWgr6riYK6AtQ","bibbaseid":"somepalli-singla-goldblum-geiping-goldstein-diffusionartordigitalforgeryinvestigatingdatareplicationindiffusionmodels-2022","author_short":["Somepalli, G.","Singla, V.","Goldblum, M.","Geiping, J.","Goldstein, T."],"bibdata":{"bibtype":"misc","type":"misc","title":"Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models","shorttitle":"Diffusion Art or Digital Forgery?","url":"http://arxiv.org/abs/2212.03860","doi":"10.48550/arXiv.2212.03860","abstract":"In this paper, Somepalli et. al. investigate the extent of digital forgery in various cutting-edge diffusion AI models, whose high quality and customizability enables them to be used for commercial art and graphic design purposes. Their findings suggest that factors such as training set size impact the rate of content replication, and that in the case of the popular Stable Diffusion model, the training data is blatantly copied to produce output. As AI art generator tools such as DALL.E and Stable Diffusion are being increasingly used for commercial purposes, the potential for AI to replicate copyrighted content without proper attribution could have significant legal implications for lawmakers, particularly as it relates to the copy-right and intellectual property regulation of AI art generators, necessitating a reevaluation of current copyright laws in the context of AI-generated art and design. (Methods and Metrics, Detecting digital (art) forgery)","urldate":"2024-01-26","publisher":"arXiv","author":[{"propositions":[],"lastnames":["Somepalli"],"firstnames":["Gowthami"],"suffixes":[]},{"propositions":[],"lastnames":["Singla"],"firstnames":["Vasu"],"suffixes":[]},{"propositions":[],"lastnames":["Goldblum"],"firstnames":["Micah"],"suffixes":[]},{"propositions":[],"lastnames":["Geiping"],"firstnames":["Jonas"],"suffixes":[]},{"propositions":[],"lastnames":["Goldstein"],"firstnames":["Tom"],"suffixes":[]}],"month":"December","year":"2022","note":"arXiv:2212.03860 [cs]","keywords":"20, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Computers and Society, Computer Science - Machine Learning, Trustworthy AI, gw_abstracts","bibtex":"@misc{somepalli_diffusion_2022,\n\ttitle = {Diffusion {Art} or {Digital} {Forgery}? {Investigating} {Data} {Replication} in {Diffusion} {Models}},\n\tshorttitle = {Diffusion {Art} or {Digital} {Forgery}?},\n\turl = {http://arxiv.org/abs/2212.03860},\n\tdoi = {10.48550/arXiv.2212.03860},\n\tabstract = {In this paper, Somepalli et. al. investigate the extent of digital forgery in various cutting-edge diffusion AI models, whose high quality and customizability enables them to be used for commercial art and graphic design purposes. Their findings suggest that factors such as training set size impact the rate of content replication, and that in the case of the popular Stable Diffusion model, the training data is blatantly copied to produce output.\n\nAs AI art generator tools such as DALL.E and Stable Diffusion are being increasingly used for commercial purposes, the potential for AI to replicate copyrighted content without proper attribution could have significant legal implications for lawmakers, particularly as it relates to the copy-right and intellectual property regulation of AI art generators, necessitating a reevaluation of current copyright laws in the context of AI-generated art and design. (Methods and Metrics, Detecting digital (art) forgery)},\n\turldate = {2024-01-26},\n\tpublisher = {arXiv},\n\tauthor = {Somepalli, Gowthami and Singla, Vasu and Goldblum, Micah and Geiping, Jonas and Goldstein, Tom},\n\tmonth = dec,\n\tyear = {2022},\n\tnote = {arXiv:2212.03860 [cs]},\n\tkeywords = {20, Computer Science - Computer Vision and Pattern Recognition, Computer Science - Computers and Society, Computer Science - Machine Learning, Trustworthy AI, gw\\_abstracts},\n}\n\n\n\n","author_short":["Somepalli, G.","Singla, V.","Goldblum, M.","Geiping, J.","Goldstein, T."],"key":"somepalli_diffusion_2022-1","id":"somepalli_diffusion_2022-1","bibbaseid":"somepalli-singla-goldblum-geiping-goldstein-diffusionartordigitalforgeryinvestigatingdatareplicationindiffusionmodels-2022","role":"author","urls":{"Paper":"http://arxiv.org/abs/2212.03860"},"keyword":["20","Computer Science - Computer Vision and Pattern Recognition","Computer Science - Computers and Society","Computer Science - Machine Learning","Trustworthy AI","gw_abstracts"],"metadata":{"authorlinks":{}},"html":""},"bibtype":"misc","biburl":"https://bibbase.org/zotero-group/dcambrid/5266609","dataSources":["e4qi3jRmPhPzc7C9a"],"keywords":["20","computer science - computer vision and pattern recognition","computer science - computers and society","computer science - machine learning","trustworthy ai","gw_abstracts"],"search_terms":["diffusion","art","digital","forgery","investigating","data","replication","diffusion","models","somepalli","singla","goldblum","geiping","goldstein"],"title":"Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Models","year":2022}