in

New Bill Demands Transparency: AI Companies Forced to Disclose Copyrighted Training Data

Hey there, my awesome followers! It’s your tech-loving funny guy, Nuked, here to share some interesting news with you. So, get ready to dive into the world of AI and copyright as we discuss a new bill that has been filed by two lawmakers.

Reps. Anna Eshoo and Don Beyer have introduced the AI Foundation Model Transparency Act. This bill aims to ensure that creators of foundation models in the field of AI disclose the sources of their training data. Why is this important? Well, it’s all about copyright protection.

Under this proposed legislation, companies developing foundation models would be required to report the origins of their training data. This way, copyright holders can have peace of mind knowing where their valuable information was taken from. The Federal Trade Commission (FTC) would collaborate with the National Institute of Standards and Technology (NIST) to establish guidelines for reporting training data transparency.

But that’s not all! The bill goes further by outlining additional requirements for AI developers. They would need to disclose how the data is retained during the inference process, describe any limitations or risks associated with the model, align the model with NIST’s AI Risk Management Framework, and provide information on the computational power used for training and running the model.

The bill also emphasizes the need for transparency when it comes to training data related to copyright. There have been several lawsuits against AI companies alleging copyright infringement, and this bill aims to address those concerns. It specifically mentions cases involving Stability AI, Midjourney, Deviant Art, and Getty Images.

The importance of accurate and unbiased information is highlighted in the bill as well. AI developers must report efforts to “red team” their models to prevent them from providing inaccurate or harmful information in various domains such as healthcare, cybersecurity, elections, finance, education, employment, public services, and vulnerable populations like children.

Now, keep in mind that this bill is still in the early stages. It needs to be assigned to a committee and undergo thorough discussions. With the upcoming election campaign season, it remains uncertain when this will happen. However, if the AI Foundation Model Transparency Act becomes law, it will establish federal rules for training data transparency.

And there you have it, my friends! An intriguing development in the world of AI and copyright. Stay tuned for more tech news and hilarious insights from yours truly, Nuked!

Spread the AI news in the universe!

What do you think?

Written by Nuked

Leave a Reply

Your email address will not be published. Required fields are marked *