A new AI law rolled out by the American House of Representatives is demanding AI firms disclose copyrighted material that’s used for training generative AI models.
The matter has been subject to much debate in the past where many companies have been wrongfully stealing data from creators without prior knowledge.
Now, an American rep has rolled out new laws that force all firms dealing in AI to showcase the copyrighted material used for training their models. While the law is yet to be passed by Congress if and when it does, it’s really going to be groundbreaking.
The law has been dubbed the Generative AI Copyright Disclosure Act forces all AI firms to be very clear about the kind of content used for training and if not, a fine would be imposed featuring $5000.
The latest law would apply to all the latest and existing models and therefore demands the submission of copyrighted material designed to create or amend training data sets of model systems as mentioned in the latest press release.
The data would be filed across the Copyright Office one month before it goes public. Moreover, it should be remembered how all such articles and posts published on social media entail a bucketload of different content. It could range from articles to posts on the app.
Therefore, it’s quite a natural and common occurrence to see the model scrape the web for a wide array of data which is needed for training generative AI. Hence, they would most likely be taking data without any sort of consent attained from the creator to whom the material belongs.
Now the fact that it has been happening for so long and not a lot of people have been taking notice is in itself a major deal for obvious reasons.
So many disagreements and disputes arose regarding this kind of material for AI model training. Did we mention how one famous case even entailed the New York Times and a legal battle speaking about tech giants Microsoft and even OpenAI?
Today, both leading AI firms are being served legal action by the New York Times which accused them of making millions by using its material without any kind of consent or compensation for training AI models.
Image: DIW-aigen
Read next: Instagram Experiments With New Blur Image Feature That Uses Machine Learning To Block Explicit Content
The matter has been subject to much debate in the past where many companies have been wrongfully stealing data from creators without prior knowledge.
Now, an American rep has rolled out new laws that force all firms dealing in AI to showcase the copyrighted material used for training their models. While the law is yet to be passed by Congress if and when it does, it’s really going to be groundbreaking.
The law has been dubbed the Generative AI Copyright Disclosure Act forces all AI firms to be very clear about the kind of content used for training and if not, a fine would be imposed featuring $5000.
The latest law would apply to all the latest and existing models and therefore demands the submission of copyrighted material designed to create or amend training data sets of model systems as mentioned in the latest press release.
The data would be filed across the Copyright Office one month before it goes public. Moreover, it should be remembered how all such articles and posts published on social media entail a bucketload of different content. It could range from articles to posts on the app.
Therefore, it’s quite a natural and common occurrence to see the model scrape the web for a wide array of data which is needed for training generative AI. Hence, they would most likely be taking data without any sort of consent attained from the creator to whom the material belongs.
Now the fact that it has been happening for so long and not a lot of people have been taking notice is in itself a major deal for obvious reasons.
So many disagreements and disputes arose regarding this kind of material for AI model training. Did we mention how one famous case even entailed the New York Times and a legal battle speaking about tech giants Microsoft and even OpenAI?
Today, both leading AI firms are being served legal action by the New York Times which accused them of making millions by using its material without any kind of consent or compensation for training AI models.
Image: DIW-aigen
Read next: Instagram Experiments With New Blur Image Feature That Uses Machine Learning To Block Explicit Content
