Good copyright office on artificial intelligence and fair use, one of the biggest questions surrounding Gen AI

If you are hoping for clarity from the United States Coptic Rights Office to train artificial intelligence – both Artificial intelligence companies Protected materials can be used for copyright under fair use, or creators can claim a violation – ready for disappointment. The office issued its third report on Friday, not the main victory companies in which it hopes to win the technology companies, nor the full that some creators sought.

The Publishing Rights Office in the United States was launched in 2023 to issue a series of reports, guidance for creators, and to deal with countless legal and ethical issues that arise from content created from artificial intelligence from programs such as ChatGPT, Gemini, Meta Ai and Dall-E. In previous reports, the copyright office has eliminated that the fully created content of artificial intelligence cannot be copied rights. Specific content It can still be eligible. These reports are not a law, but they give us a picture of how to deal with the agency to protect copyright in the era of artificial intelligence.

The third report, now available, is not the final report; It is the “Preplay” version. However, there will be no major changes in the analysis and conclusions of the Publishing Rights Office in the final report, according to what it said. WebsiteTherefore, it gives us a good understanding of the directions that it will make for future demands.

the 108 pages report In the first place, it deals with concerns related to copyrights about training artificial intelligence models-specifically, whether artificial intelligence companies have a legal submit to request an exclusion of fair use, allowing them to use copyright content without a license or compensation for copyright.

In short, the copyright office did not rule out the possibility of the case of a fair use of companies that use copyrights to train artificial intelligence. But there are some important things that would be a metaphorical blow against the issue of fair use, which shows the report in detail. Therefore, it is also possible that AI, which has been found, is also using copyright -protected materials without permission from the author, may be reasons to demand the violation of copyright. This depends on the artificial intelligence model, how it is used and what results.

The report says: “On the two ends of the spectrum, it is possible that the uses of research or non -commercial analysis that do not enable parts of works to be reproduced in the outputs.” “The other side, it is unlikely that the transcription of expressive work is unlikely to be from the sources of pirates in order to generate an unrestricted content competing in the market … it is unlikely to be fair use.

The Publishing Rights Office, which is part of the Congress Library, is the subject of the current political debate. CBS news reports The head of the department, Shira Berlmeer, known as the copyright record, was launched by President Donald Trump at the end of last week. This was a few days after Trump Firing The Secretary of the Congress Library, Carla Hayden, on Thursday. Hayden was the first African American woman and American to hold this position.

Here is what the copyright office wrote about fair use and what it needs to know about the reason for the continued growth of the legal network of the spontaneous organization and copyright.

Why fair use matters?

Technology companies are pushing strongly to obtain an exception to fair use. If they are granted such an exception, it will not matter if they have and use the work of copyright in their training data sets. The issue of potential copyright violations is in the midst of more than 30 lawsuits, including those prominent such as the New York Times against Openai and Ortiz V. Stability AI. (Disclosure: Zif Davis, the parent company CNET, filed a lawsuit against Openai, claimed that it had violated the copyright of ZifF Davis in training and operating artificial intelligence systems.)

The copyright office said in its report that these cases must continue in the judicial system: “The courts must weigh the legal factors together” due to the lack of “mechanical account or an easy formula” to determine the fair use.

Writers, actors, and other creators have pushed a strong equality with fair use. in Open message Signed by more than 400 of the largest Hollywood celebrities, creators ask the position of the Department of Science and Technology to not allow fair use. They wrote: “(America) our success stems directly from our basic respect for IP and publishing rights that are equivalent to creative risks by talented Americans and hard work from every state and lands.”

Currently, we seem to have only a few answers than before. The big questions about whether specific companies such as Openai have violated the copyright law to wait until the court is decided.

Guidelines for deciding to decide on fair use

A fair use is part of the 1976 copyright law. This ruling gives people who are not original authors the right to use copyright -protected work in specific cases, such as education, reporting and sarcasm. There are four main factors that must be taken into account in case of fair use: one, purpose of use; Two, the nature of the work; Three, the amount and tasks used; And four, its impact on the market. The report of the Publishing Rights Office analyzes the four factors in the context of artificial intelligence training.

One of the important aspects is to convert the work, which is whether AI Chatbots and the generators create outputs that are very different from the original training content. The report appears to indicate that the Chatbots AI groups used in deep search search enough. But photo generators that produce outputs in a very similar or aesthetic style may not be.

The report says that the handrails that prevent the recurrence of protected works – such as photo generators that refuse to create common slogans – will be evidence that artificial intelligence companies are trying to avoid violation. This is despite the fact that the office cited the research that proves that these handrails are not always effective, like Opeenai Ghibli Studio direction direction Show clearly.

The report argues that the content created from artificial intelligence clearly affects the market, the fourth factor. The possibility of loss of sales and marketing of markets is mentioned by excessive saturation and loss of licensing opportunities for current data markets. However, it also mentions the possibility of general benefit from developing artificial intelligence products.

The license was also highlighted, which is a common alternative to prosecuting between publishers and content owners of content, as one way possible to avoid copyright fears. Many publishers, including Financial times and Axel Springer brandsShe has made millions of dollars with artificial intelligence companies, allowing artificial intelligence developers to reach their high -quality and created content.

There are some concerns that if the license becomes the only way to achieve these data, it will give priority to large technology companies that can pay the price of this data, and the boxing of smaller developers. The Publishing Rights Office writes that these concerns should not have an impact on the analyzes of fair use and it is better to deal with them through the anti -monopoly laws and the agencies that impose it, such as the Federal Trade Committee.