The UK government’s culture secretary, Lucy Fraser, said rulemaking on AI transparency would be accelerated to help protect content creators. She said the rules would mean AI companies would have to be more transparent about the content used to train their AI models, give users the option to opt-in or opt-out of data collection for training purposes, and reward content creators whose creations are used to train models.

The Financial Times revealed the plans in an interview with Fraser. While it described the general plan, it would not share details about how rights holders can check if their material is being used in AI models.

Those close to the issue said that the government will bring proposals before the elections that should take place in the fall. This will allow stakeholders to respond to the proposals and make any suggestions or changes before they begin the process of passing the law.

The Financial Times reported that the European Union is preparing similar rules as part of its AI law. Under this law, AI developers will have to provide a sufficiently detailed summary of the content used for training and implement policies to ensure they are operating legally with EU copyright law.

AI companies like OpenAI have already predicted what governments will do and secured agreements with data providers. In recent weeks, OpenAI has announced partnerships with Reddit, Stack Overflow and the Financial Times to help it train its models.

The move will be welcomed by rights holders who claim their rights are being infringed, but for the users of these AI models, it could lead to a decrease in quality thanks to knowledge gaps. AI companies argue that they can use the data under fair use rules because their use of the data is transformative.

In the end, we will have to wait for the courts and politicians to catch up to find out whether the opinion falls on the side of rights holders or AI developers.

source: economic times