Apparently oppose the copyright Office Meta turenting
In these “general observations”, “the Copyright Office writes that” using commercial use of copyright tasks to use commercials that compete with them in existing markets, especially where it is met through illegal access, is beyond the limits of fair use. ” “
This report shows that the courts and the copyright office can also be linked to the use of pirates or illegally accessed material for AI firms’ AI training.
Judge Chbiaya preferred only Meta Torenting in the case of book authors “a kind of mess”, proper use, and the copyright office also recommended that “dataset -knowing use, which contains pyridic or illegal access tasks, should be decided against a fair use.”
However, the Copyright Office suggested that Torring should be a black mark. The Office argued that “illegal access” “tolerates” the role of use “,” despite the refusal of owners’ permission, “instead of using copyright tasks”, “pirate or illegally accessed content training goes a step further.” Perhaps if the authors could prove that the sale of pirate -related AI models was sold, the office suggested that the defense of proper use could not be avoided.
The office wrote, “The use of copyright works to create a training library, or the distribution of such library to the public will hurt the market to access these tasks.” “And where training enables a model to verbally outpit out, or to a large number of trained tasks, and those copies are easily accessible by closing users, they can substitute the sale of these tasks.”
Possibly frustrating meta -that’s Currently fighting to continue lecturing evidence In the case of the authors of the book – the Copyright Office suggests that “when the licensing is properly available,” when the licensing is properly available, copying works from marine robbers to produce unlimited content is unlikely to be eligible as proper use. “